🐛 Fix problem with node splitting in paths (#11553 )

* 🐛 Fix problem with node splitting in paths * 🐛 Fix merge splitting node in paths
🐛 Restart progressive render on canvas background change (#11643 )
2026-09-12 05:39:59 -04:00 · 2026-09-11 14:20:09 +02:00 · 2026-09-11 14:12:57 +02:00 · 2026-09-11 13:52:57 +02:00 · 2026-09-11 13:42:13 +02:00 · 2026-09-11 13:40:58 +02:00
441 changed files with 17123 additions and 5117 deletions
@@ -0,0 +1,91 @@
+# Agent skills
+
+This folder is the single home for the skills our coding agents use.
+Each skill is a folder with a `SKILL.md` inside — a short instruction
+manual that an agent loads only when it needs it.
+
+One copy serves every tool:
+
+- **opencode** reads this folder directly.
+- **Claude Code** reads it through the `.claude/skills` symlink.
+- **Codex** reads it directly.
+
+To change how the agents behave, edit the `SKILL.md` here. There is no
+second copy to keep in sync.
+
+## How the skills are organized
+
+**Flows** are the six skills you invoke by name. Each one covers one step
+in the life of a change: plan it, review the plan, implement it, review
+the code, open the pull request.
+
+**References** hold the quality standards. A flow's reviewer loads them;
+you rarely touch them directly.
+
+**Procedures** define how one concrete step is done — a plan document, an
+issue, a commit. Flows call them, but they also work on their own.
+
+**Utilities** are small helpers for everyday work: search, file lookup,
+JSON, REPL access, and so on.
+
+## Flows
+
+| Skill | What it does | When you would say |
+|---|---|---|
+| [`make-a-plan`](skills/make-a-plan/SKILL.md) | Researches the task, writes an implementation plan, asks you the open questions in plain language, and saves the plan to `.agents/plans/`. | "make a plan for the token refresh bug" |
+| [`review-plan`](skills/review-plan/SKILL.md) | Evaluates a plan before anyone writes code: completeness, ordering, risks. Approves it or asks for changes. | "review this plan before we start" |
+| [`implement-plan`](skills/implement-plan/SKILL.md) | Shows you the full flow first — the issue and branch it will create (or the branch it continues on), the execution style, and the task checklist — and, after your go-ahead, executes a ready plan. Default: every task, one commit. On request ("step by step"): one task, one commit, your confirmation between tasks. On request ("direct"): no issue and no branch, commits on the current branch. | "implement the plan" · "step by step, one commit per task" · "direct, no branch" |
+| [`review-code`](skills/review-code/SKILL.md) | Reviews a diff, branch, or PR and returns findings ranked by impact. | "review my changes before I push" |
+| [`create-pr`](skills/create-pr/SKILL.md) | Opens a pull request for the current branch — with checks on base branch, commits, issue, and push state — or updates an existing PR's title and description. | "open a PR for this branch" |
+| [`resolve-git-conflicts`](skills/resolve-git-conflicts/SKILL.md) | Untangles merge or rebase conflicts: explains both sides, proposes a resolution, applies it after you approve. Never runs `git rebase --continue`. | "resolve these conflicts" |
+
+## References
+
+| Skill | What it holds |
+|---|---|
+| [`plan-review-criteria`](skills/plan-review-criteria/SKILL.md) | The plan review rubric: six axes, severity levels, approval standard, output format. The `review-plan` reviewer loads it. |
+| [`code-review-criteria`](skills/code-review-criteria/SKILL.md) | The code review rubric: five axes, core principles (DRY, KISS, YAGNI), severity format, verdict. The `review-code` reviewer loads it. |
+
+## Procedures
+
+| Skill | What it does |
+|---|---|
+| [`planner`](skills/planner/SKILL.md) | The spec of a good plan: context, architecture decisions, tasks with acceptance criteria, checkpoints. Used by `make-a-plan`. |
+| [`create-issue`](skills/create-issue/SKILL.md) | Creates a GitHub issue that follows Penpot conventions. Used by `implement-plan`; also works on its own. |
+| [`create-commit`](skills/create-commit/SKILL.md) | Makes a commit the Penpot way: emoji subject, clear body, `AI-assisted-by` trailer. Used by `implement-plan`; also works alone when you say "commit this". |
+
+## Utilities
+
+| Skill | What it does |
+|---|---|
+| [`bat-cat`](skills/bat-cat/SKILL.md) | Read files in the terminal with syntax highlighting and line numbers. |
+| [`fd-find`](skills/fd-find/SKILL.md) | Find files by name or pattern, respecting `.gitignore`. |
+| [`ripgrep`](skills/ripgrep/SKILL.md) | Fast content search with regular expressions. |
+| [`jq-json-processor`](skills/jq-json-processor/SKILL.md) | Slice, filter, and reshape JSON output. |
+| [`nrepl-eval`](skills/nrepl-eval/SKILL.md) | Run Clojure or ClojureScript code in the live REPL sessions (backend and frontend). |
+| [`taiga`](skills/taiga/SKILL.md) | Look up Penpot issues, user stories, and tasks in Taiga. |
+| [`testing`](skills/testing/SKILL.md) | The repo's testing rules and TDD workflow, loaded before writing tests. |
+| [`local-ci`](skills/local-ci/SKILL.md) | Run CI-style lint, test, and format checks for the modules you touched with `scripts/ci`, and read the logs when they fail. |
+| [`security-and-hardening`](skills/security-and-hardening/SKILL.md) | Security checks for code that handles user input, auth, or external services. |
+| [`ste`](skills/ste/SKILL.md) | Rewrites prose in Simplified Technical English. Loads only when you name it. |
+| [`refine-prompt`](skills/refine-prompt/SKILL.md) | Rewrites a rough prompt into a clearer one. Never runs the prompt. |
+| [`update-changelog`](skills/update-changelog/SKILL.md) | Regenerates `CHANGES.md` from a GitHub milestone. |
+
+## A typical round
+
+1. `/make-a-plan` — you get a plan and a saved file in `.agents/plans/`.
+2. `/review-plan` — a second opinion; approve or request changes.
+3. `/implement-plan` — the code gets written and committed. Starting from a base branch, it also opens the GitHub issue and the `issue-NNNN` branch; the plans that follow continue on that same branch.
+4. `/review-code` — a reviewer checks the commit.
+5. `/create-pr` — the branch goes up as a pull request.
+
+Every step also works on its own, and you can always say what you want
+in plain words — the agents pick the right skill from what you say.
+
+## Adding or changing a skill
+
+Create a folder here with a `SKILL.md` inside. The file needs `name` and
+`description` in its frontmatter, and a clear "When to use" section so
+agents know when to reach for it. Keep one job per skill, and keep the
+two families apart: flows are named with a verb first; reference skills
+end in `-criteria`.
@@ -9,6 +9,11 @@ metadata: {"clawdbot":{"emoji":"🦇","requires":{"bins":["bat"]},"install":[{"i

 `cat` with syntax highlighting, line numbers, and Git integration.

+## When to use
+
+- Reading or displaying a file in the terminal — prefer it over plain
+  `cat`: syntax highlighting, line numbers, git-side indicators.
+
 ## Quick Start

 ### Basic usage
@@ -1,9 +1,9 @@
 ---
-name: code-review
-description: Conducts multi-axis code review. Use before merging any change. Use when reviewing code written by yourself, another agent, or a human. Use when you need to assess code quality across multiple dimensions before it enters the main branch.
+name: code-review-criteria
+description: Code review criteria — the five review axes, core principles, severity format, and verdict for reviewing code changes. Loaded by the reviewer subagent of the review-code flow. Not a user-facing flow — to review code, use the review-code flow.
 ---

-# Code Review and Quality
+# Code Review Criteria and Quality

 ## Overview

@@ -13,11 +13,10 @@ Multi-dimensional code review with quality gates. Every change gets reviewed bef

 ## When to Use

- Before merging any PR or change
- After completing a feature implementation
- When another agent or model produced code you need to evaluate
- When refactoring existing code
- After any bug fix (review both the fix and the regression test)
+- The reviewer subagent of the `review-code` flow loads this skill to perform
+  the review of a code change.
+- To review code, always go through the `review-code` flow — never load this
+  skill directly for that. This is the criteria reference, not the flow.

 ## Core Principles

@@ -20,6 +20,17 @@ Before drafting any commit, read `mem:workflow/creating-commits` end-to-end. It
 is the authoritative source for the commit message format, the emoji menu,
 subject/body limits, and the `AI-assisted-by` trailer. Follow it exactly.

+## Iron Rules (non-negotiable)
+
+1. **Wrap every body line at 76 characters or fewer.** Count characters, do
+   not eyeball. Exceptions: `Signed-off-by:` / `AI-assisted-by:` trailers and
+   lines carrying a URL. This is the rule agents skip most often.
+2. **Subject ≤70 chars**, imperative, capitalized, no trailing period.
+3. **Blank line between subject and body.**
+4. **Run `./scripts/check-commit` and require exit code 0.** It mechanically
+   checks rules 1–3. A non-zero exit is a hard blocker: fix the message and
+   re-commit. Never report the commit as done with a failing checker.
+
 ## Workflow

 1. **Stage the files** specified by the calling context. Do not ask for
@@ -29,12 +40,18 @@ subject/body limits, and the `AI-assisted-by` trailer. Follow it exactly.
   that does not match the stated intent, **STOP** and tell the user before
   committing.
 3. Draft the message following the format in the memory doc, wrapping the body
-   at 72 characters per line, and run:
+   at 76 characters per line, and run:
   ```bash
   git commit -m "<subject>" -m "<body>"
   ```
   (or `git commit -F -` if the body has unusual characters).
-4. The `AI-assisted-by` trailer value is provided by the calling context — use
+4. **Verify the message with the checker**:
+   ```bash
+   ./scripts/check-commit
+   ```
+   If it fails, amend the message (`git commit --amend`) until it passes. Do
+   not finish with a failing checker.
+5. The `AI-assisted-by` trailer value is provided by the calling context — use
   it verbatim.

 ## Constraints
@@ -45,3 +62,4 @@ subject/body limits, and the `AI-assisted-by` trailer. Follow it exactly.
 - Do not amend a commit you did not create in this session, unless explicitly asked.
 - Do not bypass pre-commit hooks (`--no-verify`) unless explicitly asked.
 - Do not add untracked files that were not created in this session.
+- Do not skip the `scripts/check-commit` verification step (Iron Rule 4).
@@ -0,0 +1,105 @@
+---
+name: create-pr
+description: PR flow — open a new PR for the current task branch (validates base branch, commits, issue and push state) or update an existing PR's title or description to match Penpot conventions. Use it when the user asks to open or create a PR, in any phrasing.
+---
+
+# Create PR
+
+Two modes. **Open mode** takes the current task branch to a new, validated
+PR. **Update mode** rewrites an existing PR's title or description. Gather
+information, validate, and act in one pass. If validation fails, STOP with a
+single coherent message that lists every problem and states exactly what
+information is missing — never fix or work around problems silently.
+
+Both modes require an authenticated `gh` CLI (`gh auth status`) and never
+push — the user pushes from their own shell.
+
+## When to use
+
+- The user asks to open or create a NEW PR for the current task branch, in
+  any phrasing ("open a PR", "create the pull request", "put this up for
+  review") — or runs `/create-pr`. → **Open mode**.
+- The user asks to fix or update an EXISTING PR's title or description to
+  match conventions. → **Update mode**.
+
+If the running agent cannot write (for example, the plan agent), say so and
+stop — this skill needs the build agent.
+
+## Open mode
+
+### 1. Gather context (read-only)
+
+- Current branch: `git rev-parse --abbrev-ref HEAD`.
+- Target base branch: run `./scripts/detect-target-branch` from the repo root.
+  It prints the nearest ancestor branch of HEAD (exit 0) or fails (exit 1).
+- Commits: `git log --oneline <base>..HEAD`.
+- Push state (local): `git rev-parse --verify origin/<branch>` and compare
+  with HEAD. It reads the local remote-tracking ref — no network, no SSH. It
+  reflects the last push or fetch this clone knows about.
+- Issue: from the session context, or from the branch name — `issue-NNNN`
+  maps to issue NNNN; recover its title and body with `gh issue view NNNN`.
+
+### 2. Validate — stop with one message if anything fails
+
+Run all checks before reporting, then report every failure together:
+
+1. **Base branch not usable.** If the script fails (exit 1), or its output —
+   after stripping an optional `remotes/origin/` prefix — is not one of the
+   canonical branches (`develop`, `staging`, `main`), stop and ask the user
+   to re-run with more context — for example, passing the base branch
+   explicitly in their invocation. An explicit base given by the user
+   overrides the script's output.
+2. **On a base branch.** There is no task branch to merge — say so and stop.
+3. **No commits.** The branch has no commits ahead of the base — say so and
+   stop.
+4. **No clear issue.** There is no issue in the session context, and the
+   branch name has no `issue-NNNN` pattern (or `gh issue view` finds nothing)
+   — say so and stop. Exception: the user's invocation says `no issue` /
+   `without issue` — then continue without an issue reference.
+5. **Branch not pushed.** The remote-tracking ref `origin/<branch>` is
+   missing, or `git rev-parse origin/<branch>` differs from HEAD — the
+   branch was never pushed, or has commits the remote does not have. Never
+   push yourself; ask the user to push and to run `/create-pr` again
+   afterwards, then stop.
+
+### 3. Already-open PR
+
+Check whether a PR already exists for this branch (`gh pr list --head
+<branch>`). If one exists, report its URL and stop — do not create a second
+one. Title or description fixes belong to Update mode.
+
+### 4. Write and create the PR
+
+Write the title and body following `mem:workflow/creating-prs` (title format,
+description structure, writing principles) and `mem:workflow/creating-commits`
+(commit type emojis). Derive the title and body from the commits and, when
+there is one, from the issue body. Reference the issue with `Closes #NNNN`.
+
+```bash
+gh pr create --repo penpot/penpot --title "<TITLE>" --body-file /tmp/pr-body.md
+```
+
+### 5. Report
+
+Report the PR URL and stop.
+
+## Update mode
+
+1. Identify the PR: the number given by the user, or `gh pr list --head
+   <branch>`.
+2. Write the new title and/or body following `mem:workflow/creating-prs`.
+3. Apply and verify:
+
+```bash
+gh pr edit <NUMBER> --repo penpot/penpot --title "<TITLE>" --body-file /tmp/pr-body.md
+gh pr view <NUMBER> --repo penpot/penpot --json title,body
+```
+
+4. Report and stop.
+
+## User context
+
+Extra context in the user's invocation (the message that triggered this skill)
+plays the role command arguments play elsewhere: overrides such as `no issue` /
+`without issue`, an explicit base branch (`from origin/staging`), a PR number
+for Update mode, and so on.
@@ -9,6 +9,11 @@ metadata: {"clawdbot":{"emoji":"📂","requires":{"bins":["fd"]},"install":[{"id

 User-friendly alternative to `find` with smart defaults.

+## When to use
+
+- Locating files or directories by name or pattern — prefer it over
+  plain `find`: simpler syntax, smart defaults, respects `.gitignore`.
+
 ## Quick Start

 ### Basic search
@@ -0,0 +1,144 @@
+---
+name: implement-plan
+description: Implementation flow — execute a ready plan from the session context: read the plan, detect the flow, then present the full picture (issue and branch to create or the branch to continue on, execution style, task checklist) and wait for confirmation. Default is every task with one final commit; on request ("step by step"), one task and one commit at a time with a pause after each; on request ("direct"), no issue and no branch — the commit lands on the current branch. Use it when the user asks to implement or execute a plan, in any phrasing.
+---
+
+# Implement Plan
+
+This flow is run once a plan is ready (for example, from plan mode). Execute
+the plan already prepared in the current session context. It never pushes —
+the user pushes.
+
+By default it ends with exactly one commit. When the user asks for it
+("step by step"), it commits once per task instead and waits for the
+user's confirmation after each one (see *Execution modes*).
+
+## When to use
+
+- The user asks to implement or execute a plan, in any phrasing:
+  "implement the plan", "execute it", "go build it" — or runs
+  `/implement-plan`.
+- A ready, reviewed plan is in the session context or a plan file path
+  was given (typically after `/make-a-plan` or `/review-plan`).
+
+Do not use it to produce plans — that is the `make-a-plan` flow.
+
+## 1. Read the plan first
+
+Identify the plan to execute — from the file path the user gave, the
+arguments, or the session context. Read it completely. Read the required
+memories before writing any code: `mem:critical-info` and the core memory
+of every module the plan touches, plus the deeper memories they reference
+(AGENTS.md governs this).
+
+## 2. Detect the flow (no questions)
+
+Inspect the current branch with `git rev-parse --abbrev-ref HEAD`, pick the
+mode, and announce it in one line before presenting anything. Detection is
+read-only: nothing is created until the user confirms (step 3).
+
+- **On a base branch** (`main`, `develop`, `staging`) → **standalone mode**:
+  a new GitHub issue and a branch `issue-NNNN` will be created after the
+  user's confirmation.
+- **On any other branch** (a feature branch, typically `issue-NNNN`) →
+  **continue mode**: the implementation continues on the current branch.
+  No issue or branch is created. The branch name provides the issue
+  reference when it follows the `issue-NNNN` pattern.
+
+Arguments override detection: `standalone`, `continue`, `direct`
+(`no branch` / `direct commit`), `no issue` / `without issue`, or an
+explicit base such as `from origin/develop`.
+
+**Direct mode** (`direct`, `no branch`, `direct commit`): no issue and
+no branch — the implementation and the commit land on the current branch
+as it is, even when it is a base branch. Best for small or tooling-only
+changes the user wants committed in place.
+
+**Standalone while already on a feature branch:** stop and explain that this
+would stack branches. Ask the user to re-run with an explicit base, for
+example `from origin/develop` — then branch from that base instead of HEAD.
+
+## 3. Present the checklist and wait
+
+Before touching the repository, show the user the full picture:
+
+- **The flow**: whether the GitHub issue and the branch will be created
+  (standalone mode — give the planned branch name, `issue-NNNN` or
+  `plan-<slug>`), whether you continue on the current branch
+  (continue mode — name it), or whether everything lands on the current
+  branch as it is (direct mode — name it, and say so when it is a base
+  branch).
+- **The execution style**: batch or step-by-step (see *Execution modes*).
+- A checklist (todolist) of the plan's tasks, in order.
+
+Then WAIT for the user's explicit confirmation. Do not start until you
+have it. If the plan has no discrete tasks, ask the user how to split
+it, or propose running it as a single change.
+
+## 4. Execute the plan
+
+**Standalone setup, after the confirmation:** create the issue with the
+**`create-issue`** skill, following the *Creating Issues from Draft Body*
+flow in `mem:workflow/creating-issues`. Derive the issue title and body
+from the plan, capture the new issue's number — call it **NNNN** — and
+create the branch from the current HEAD:
+
+```
+git checkout -b issue-NNNN
+```
+
+If the arguments say `no issue` / `without issue`, skip the issue and
+create a branch named `plan-<slug>` instead, where `<slug>` is the plan
+title, lowercase and hyphen-separated.
+
+If the arguments say `direct` / `no branch` / `direct commit`, skip the
+issue and the branch: implement and commit on the current branch as it
+is. If it is a base branch, the checklist presentation already said so —
+no further confirmation is needed.
+
+### Batch mode (default)
+
+Implement every task in one go. Work methodically, keeping changes
+focused on what the issue requires. Respect the plan's proposed
+parallelization when it applies.
+
+When the implementation is complete, load the **`create-commit`** skill
+and follow its workflow to commit the changes. Provide a brief summary
+of what was implemented and why, the issue reference (`issue-NNNN`) when
+there is one, and the model name you are running as so the
+`AI-assisted-by` trailer is set correctly.
+
+### Step-by-step mode (on request)
+
+When the user asks for it — "step by step", "task by task", "one commit
+per task" — loop one task at a time:
+
+- Execute exactly ONE task.
+- Commit it now: load the **`create-commit`** skill and follow it —
+  one commit per task, never two tasks in one commit. Same inputs as
+  always: what and why, the issue reference, your model name.
+- Show the user the result (what changed, files touched, how it was
+  verified).
+- WAIT for the user's confirmation before starting the next task.
+
+Never batch in this mode: no two tasks in one commit, and no new task
+before the user confirms. If a task turns out much bigger than planned,
+stop and ask the user before splitting it.
+
+## When you are done
+
+End by suggesting the next steps (suggestions, not a required pipeline — any
+instruction from me overrides them):
+
+- `/review-code` — to review the changes just committed; it routes to
+  `/make-a-plan` by itself if the findings need one.
+- `/create-pr` — when the task is done and the branch is ready to merge.
+
+## User context
+
+Extra context in the user's invocation (the message that triggered this
+skill) plays the role command arguments play elsewhere: `standalone`,
+`continue`, `direct` (`no branch` / `direct commit`), `no issue` /
+`without issue`, an explicit base such as `from origin/develop`, or
+`step by step` / `one commit per task` for the step-by-step execution
+mode. Modes combine freely, for example "standalone step by step".
@@ -9,6 +9,11 @@ metadata: {"clawdbot":{"emoji":"🔍","requires":{"bins":["jq"]},"install":[{"id

 Process, filter, and transform JSON data with jq.

+## When to use
+
+- Parsing, filtering, or transforming JSON from commands, files, or API
+  responses — slicing, reshaping, or validating JSON output.
+
 ## Quick Examples

 ### Basic filtering
@@ -0,0 +1,95 @@
+---
+name: local-ci
+description: Run local CI-style checks with ./scripts/ci (lint, tests, format) per monorepo module. Use when verifying changes before declaring work done, running lint or tests locally, fixing formatting, or repairing Clojure delimiter errors.
+---
+
+# Local CI
+
+Run the same checks CI runs, locally, for the modules you touched, with
+`scripts/ci`. Each task writes a log file; the final summary says what
+passed and what failed.
+
+Full details: `mem:scripts/ci` (file: `.serena/memories/scripts/ci.md`)
+
+## When to use
+
+- After implementing or fixing code — verify every module you touched
+  before declaring the work done.
+- When the user asks to run CI, lint, tests, or format checks locally.
+- When you changed `common/` — validate its consumers too.
+
+**Skip:** while exploring, planning, or reading code.
+
+## Command reference
+
+Run from the repo root:
+
+```bash
+./scripts/ci [OPTIONS] [MODULES...]
+```
+
+Modules: `frontend` `backend` `common` `render-wasm` `exporter` `mcp`
+`plugins` `library`, or `--all` for every module.
+
+With no task flags it runs three tasks per module, in order: **lint**,
+**test**, **fmt** (format check; `--fix` formats files instead).
+
+| Flag | Effect |
+|------|--------|
+| `--all` | Run every module |
+| `--exclude MOD` | Skip one module (repeatable) |
+| `--lint` / `--no-lint` | Run only lint / drop lint |
+| `--test` / `--no-test` | Run only tests / drop tests |
+| `--fmt` / `--no-fmt` | Run only format check / drop it |
+| `--fix` | Format files instead of checking (other tasks unaffected) |
+| `--paren-repair` | Fix delimiter errors in Clojure/CLJS files |
+| `--fail-fast` | Stop at the first failure |
+| `--quiet` | Suppress failure output |
+| `--dry-run` | Show what would run, execute nothing |
+| `--clean` | Delete the `.ci-logs/` directory |
+
+## Reading failures
+
+Every task writes its full output to `.ci-logs/<module>-<task>.log`. On
+failure the script prints only the last 30 lines. To diagnose a failure,
+**read the log file** — never re-run the command piped through filters
+(repo rule: redirect to a file first, then read it). The exit code is 1
+when any task failed; the summary lists each failed `module:task` and its
+log path.
+
+## Typical workflows
+
+```bash
+# Verify a module you changed: lint + tests + format check
+./scripts/ci frontend
+
+# Fast pass while iterating: lint only
+./scripts/ci --lint frontend
+
+# Lint + format check, skip the long test suite
+./scripts/ci --no-test frontend
+
+# Format the module without running the test suite
+./scripts/ci --fix --no-test frontend
+
+# Broke delimiters in Clojure/CLJS files: repair first, then lint
+./scripts/ci --paren-repair frontend
+./scripts/ci --lint frontend
+
+# Changed common/ — validate its consumers too
+./scripts/ci frontend backend exporter
+
+# Preview what would run, without running it
+./scripts/ci --dry-run --all
+```
+
+## Gotchas
+
+- Run from the repo root.
+- Test tasks are long-running (backend runs `clojure -M:dev:test`); give
+  the bash call a generous timeout (10–20 minutes) instead of letting it
+  time out mid-run.
+- `mcp` has no lint task — it shows as skipped, not failed.
+- `--paren-repair` only fixes delimiters; run lint afterwards to catch
+  what remains. See `mem:scripts/paren-repair`.
+- What to run and how to read test results: `mem:testing`.
@@ -0,0 +1,100 @@
+---
+name: make-a-plan
+description: Planning flow — research the subject of this session, produce an implementation plan with the planner skill, resolve open questions with the user in plain language, and save the final plan to .agents/plans/. Use it when the user asks to plan, design, or break down a task, in any phrasing.
+---
+
+# Make a Plan
+
+Act as a senior software engineer: research the subject of this session in depth and
+produce a well-grounded, actionable implementation plan.
+
+If the running agent cannot write (for example, the plan agent), say so and
+stop — this skill needs the build agent to save the plan.
+
+## When to use
+
+- The user asks to plan, design, or break down a task, in any phrasing:
+  "make a plan", "how would we build X", "design an approach for Y" —
+  or runs `/make-a-plan`.
+- The user asks to rework or extend an existing plan (for example, after
+  review findings) — revise the saved plan file in place.
+
+Do not use it to execute a plan — that is the `implement-plan` flow.
+
+## Instructions
+
+1. **Produce the plan** with the `planner` skill. By default, research the
+   subject of this session and draft the plan yourself. If I ask for it (for
+   example, `delegated` in the user context), delegate to the `general` subagent
+   instead — the delegate must also follow the `planner` skill and receive all
+   the relevant session context (a review, user feedback, and so on).
+2. Before asking me to decide anything, explain the plan and every open question in
+   plain language. Assume I know only the high-level project goal, not the codebase,
+   architecture, implementation terms, or the problem this task solves.
+3. Once all decisions are answered and the plan is final, save it verbatim to the
+   announced path under `.agents/plans/` (create the directory if it does not
+   exist). This step is the flow's explicit authorization to write the plan
+   file — the only write allowed here. If I later ask for changes, update the
+   saved file directly.
+4. Present me with a clear, self-contained summary of the plan's most relevant points
+   only after all required decisions have been answered. Write it for someone who knows
+   only the project's high-level goal and may not know the plan's low-level context.
+   Explain necessary technical language in plain terms, include the problem being
+   solved and the proposed outcome, and do not assume that listing technical task names
+   is enough.
+
+### Hard rule — read-only while planning
+
+While this flow runs, act read-only: research with read-only tools only.
+Never edit source files, never run builds, tests, linters, or any command that
+modifies state, and never commit. The single allowed write is the plan file in
+step 3. This rule expires when I approve the plan or move on to another task;
+then you act as a normal build agent again.
+
+When the plan contains open questions, do not show them as bare technical questions or
+assume that I understand the technical language or technical words used in the plan.
+For each question, first explain:
+
+- What part of the user problem the decision affects.
+- The relevant concept from the beginning, with a small concrete example.
+- What each available option would make the system do.
+- The practical benefits, costs, risks, and user-visible consequences of each option.
+- Which option the planner recommends and why.
+
+Only after that explanation, use the `question` tool to ask the decision with clear,
+non-technical option labels. Put the recommended option first and mark it as
+`(Recommended)`. Group related questions when their context is shared, but do not ask a
+question whose meaning has not already been explained.
+
+If I say that I do not understand a question or its choices, do not treat my previous
+answer as valid. Explain the concepts again from the high-level project goal, use a more
+concrete example, explain the implications, and ask the question again with the
+`question` tool. Repeat this until I can make an informed choice. If one answer creates
+new design consequences or additional decisions, explain those consequences before
+asking any new question.
+
+Distinguish clearly between requirements already fixed by the roadmap or existing
+architecture and choices that actually require my input. Do not ask me to choose an
+implementation detail when the plan can resolve it safely without changing the public
+behavior. If there are no decisions that require my input, say so and present the
+summary.
+
+IMPORTANT: **Under no circumstances execute the plan. Wait for the user to review it
+after all possible questions have been answered.** The final summary must explain the
+problem being solved, the proposed behavior, the main user-visible workflow, important
+constraints and risks, what is deliberately out of scope, and the path where the plan
+is saved. Never assume that a short list of task names is enough context. End
+the final response by suggesting the next steps, in this order:
+
+1. `/review-plan` — to get a second opinion on the plan before executing it.
+2. `/implement-plan` — to execute the plan from the current session context.
+
+These are suggestions, not a required pipeline — any instruction from me
+overrides them (for example, asking you to implement the plan directly).
+
+## User context
+
+Extra context in the user's invocation (the message that triggered this skill)
+plays the role command arguments play elsewhere: for example, `delegated` to
+hand the research and drafting to the `general` subagent, or corrections and
+feedback about a previous plan.
@@ -10,6 +10,12 @@ Evaluate Clojure (or ClojureScript) code via a running nREPL server using

 Full documentation: `mem:scripts/nrepl-eval` (file: `.serena/memories/scripts/nrepl-eval.md`)

+## When to use
+
+- Evaluating Clojure or ClojureScript code against the running nREPL
+  sessions (backend 6064, frontend 3447) — live inspection, patching, or
+  debugging.
+
 ## Quick Reference

 ```bash
@@ -1,9 +1,9 @@
 ---
-name: plan-review
-description: Reviews implementation plans for quality, completeness, and actionability. Use after a plan is produced by the planner skill, before starting implementation. Use when evaluating a plan written by yourself, another agent, or a human.
+name: plan-review-criteria
+description: Plan review criteria — the six review axes, severity rubric, approval standard, and output format for reviewing implementation plans. Loaded by the reviewer subagent of the review-plan flow. Not a user-facing flow — to review a plan, use the review-plan flow.
 ---

-# Plan Review
+# Plan Review Criteria

 ## Overview

@@ -13,10 +13,10 @@ Multi-dimensional plan review with quality gates. Every plan gets reviewed befor

 ## When to Use

- After the planner skill produces a plan
- Before starting implementation on any non-trivial task
- When reviewing a plan written by another agent or a human
- When a plan feels too large, vague, or risky to start
+- The reviewer subagent of the `review-plan` flow loads this skill to perform
+  the review of a plan.
+- To review a plan, always go through the `review-plan` flow — never load this
+  skill directly for that. This is the criteria reference, not the flow.

 **Do NOT use for:** Single-file changes with obvious scope, or when the task is trivial enough to just do.

@@ -87,7 +87,7 @@ Can an implementer actually execute this?

 ### 6. Proposed Code Quality *(when the plan includes implementation details)*

-If the plan proposes code shapes, function signatures, data structures, or API designs, evaluate those proposals against `code-review` criteria:
+If the plan proposes code shapes, function signatures, data structures, or API designs, evaluate those proposals against `code-review-criteria`:

 - **Correctness:** Do the proposed types/signatures handle edge cases (null, empty, boundaries)?
 - **Readability:** Are proposed names descriptive and consistent with project conventions?
@@ -215,7 +215,7 @@ Check that the plan can actually confirm it worked:
 If the plan includes code snippets, types, or API designs:

 ```
- Load code-review skill for criteria
+- Load code-review-criteria skill for criteria
 - Check proposed signatures for edge cases
 - Verify naming follows project conventions
 - Confirm abstractions follow existing patterns
@@ -310,6 +310,6 @@ If the plan includes code snippets, types, or API designs:
 ## See Also

 - For producing plans, use the `planner` skill
- For reviewing implemented code, use `code-review` — also the criteria source for axis 6
+- For reviewing implemented code, use `code-review-criteria` — also the criteria source for axis 6
 - For security-specific concerns, see `security-and-hardening`
 - For testing strategy guidance, see `testing`
@@ -0,0 +1,205 @@
+---
+name: planner
+description: Read-only planning and architecture analysis — produce a structured implementation plan with task breakdown, acceptance criteria, sizing, and checkpoints. Always output to the user with the plan, suggested save path and the next steps.
+---
+
+# Planner
+
+Produce a plan that another engineer or agent can execute without guessing.
+
+## When to Use
+
+- The user asks for a plan, design, or analysis of a feature or bug.
+- The user wants to understand which parts of the codebase a task will touch.
+- The user needs a step-by-step implementation plan with file paths, function
+  names, and test strategy.
+- The user asks "how would I implement X?" or "what's involved in fixing Y?".
+- The user is about to start non-trivial work and wants a bite-sized task
+  breakdown.
+- A task feels too large or vague to start.
+- Work needs to be parallelized across multiple agents or sessions.
+
+Do not use for a small change with obvious scope or an existing executable plan.
+
+## CRITICAL: Required Reading Before Planning
+
+Before drafting any plan, work through the project's own guidance:
+
+1. Read `critical-info` (`.serena/memories/critical-info.md`) — the entry point
+   that describes the monorepo structure and module dependency graph.
+2. From `critical-info`, identify which modules your task affects.
+3. Read each affected module's core memory, e.g. `mem:frontend/core`,
+   `mem:backend/core`, `mem:common/core`, `mem:exporter/core`,
+   `mem:render-wasm/core`. Follow `mem:` references deeper as needed.
+4. For each affected module, note its lint, format, and test commands so the
+   plan can include concrete verification steps.
+
+Skipping this step is the #1 cause of incorrect or incomplete plans.
+
+## Constraints
+
+- You are **analysis-only** — never create, edit, or delete source code. The
+  only file you may write is the plan itself, and only when the command or
+  user explicitly instructs you to save it.
+- You do **not** run builds, tests, linters, or any commands that modify state.
+- You do **not** create git commits or interact with version control.
+- You do **not** execute shell commands beyond read-only searches (`rg`, `ls`,
+  `find`, `cat`, `bat`).
+- Your output is a structured plan or analysis, ready for handoff to an
+  engineer agent or developer.
+
+## Planning Process
+
+1. Define the problem, desired outcome, constraints, and exclusions.
+2. Trace the current behavior through the affected modules.
+3. Map dependencies and choose an implementation order that builds foundations
+   before their consumers.
+4. Identify open product or architecture decisions. Resolve implementation
+   details from existing conventions when they do not affect public behavior.
+5. Identify edge cases, security and data risks, performance bounds, breaking
+   changes, and external dependencies.
+6. Split the work into small, ordered tasks. Prefer complete testable slices
+   over unrelated layer-wide batches. Apply DRY and KISS to the proposed
+   implementation.
+7. Define exact acceptance criteria and verification for every task.
+8. Add a checkpoint after every two or three tasks in a longer plan.
+9. State which tasks can run in parallel and which must remain sequential.
+
+## Task Format
+
+Each task follows this structure:
+
+```markdown
+## Task [N]: [Short descriptive title]
+
+**Description:** One or two paragraphs explaining what this task accomplishes.
+Should be clear and concise.
+
+**Rationale:** Why this task exists and why this approach over the obvious
+alternatives — design decisions, trade-offs, constraints discovered during
+analysis. One or two sentences; skip only if genuinely trivial.
+
+**Code sketch (optional):** Signature-, type-, or shape-level example when the
+intended interface is non-obvious. Keep it short — a skeleton that fixes the
+contract (function signature, model fields, error shape), never a full
+implementation. Omit when the task is mechanical.
+
+**Acceptance criteria:**
+- [ ] [Specific, testable condition]
+- [ ] [Specific, testable condition]
+
+**Verification:**
+- [ ] Relevant tests pass (module-specific test command).
+- [ ] Lint/formatter passes (module-specific check command), if applicable.
+- [ ] The core flow works end-to-end, if applicable.
+
+**Dependencies:** [Task numbers this depends on, or "None"]
+
+**Files likely touched:**
+- `path/to/file.clj`
+- `path/to/file_test.clj`
+
+**Estimated scope:** [XS: 1 file | S: 1-2 files | M: 3-5 files | L: 5+ files]
+```
+
+Use commands from `mem:testing` and affected module memories. Never substitute
+generic text such as "run the tests" when the project documents an exact
+command.
+
+When possible, design each task with TDD in mind: acceptance criteria double as a test
+list, and the natural first step of the task is writing those tests before the
+implementation. Some tasks resist this (config, migrations, pure wiring) — for those, keep
+the usual verification steps.
+
+## Task Sizing
+
+| Size | Files | Scope | Example |
+|------|-------|-------|---------|
+| **XS** | 1 | Single function, config change, or schema tweak | Add a validation rule |
+| **S** | 1-2 | One handler or component method | Add a new RPC endpoint |
+| **M** | 3-5 | One vertical feature slice | Bookmark CRUD with tests |
+| **L** | 5-8 | Multi-component feature | Search with filtering and pagination |
+| **XL** | 8+ | **Too large — break it down further** | — |
+
+Split a task when it contains independent outcomes, spans unrelated systems, or cannot be
+completed and verified in one focused session (if a task is XL, it should be broken into
+smaller tasks; agents perform best on S and M tasks).
+
+## Task order and checkpoints
+
+Arrange tasks so that:
+
+1. Dependencies are satisfied (build foundation first)
+2. Each task leaves the system in a working state
+3. Verification checkpoints occur after every 2-3 tasks
+4. High-risk tasks are early (fail fast)
+
+Add explicit checkpoints with the relevant module commands:
+
+```markdown
+### Checkpoint: After Tasks 1-3
+- [ ] Relevant tests pass (module-specific command).
+- [ ] The relevant build or compilation passes, if applicable.
+- [ ] The core flow works end-to-end.
+```
+
+## Output Format
+
+The plan is always delivered in the response so the user sees it regardless
+of which agent is running the skill. File writes follow `Constraints` —
+by default announce the path instead of writing.
+
+Announce the save path `.agents/plans/YYYY-MM-DD-<slug>.md` (today's date,
+lowercase hyphen-separated slug, e.g. `2026-09-10-add-batch-get-profiles`;
+an explicit user path wins).
+
+End the response by suggesting the next steps: `/review-plan` to get a second
+opinion on the plan and `/implement-plan` to execute it.
+
+### Plan Structure
+
+Use this document shape:
+
+```markdown
+# Plan: Title
+
+## Context
+## Affected Modules
+## Architecture Decisions
+## Risks and Considerations
+## Approach
+## Task List
+## Verification and Testing
+## Parallelization
+## Open Questions
+```
+
+Omit empty sections only when they do not apply. Every implementation task
+still requires acceptance criteria, verification, dependencies, likely files,
+and scope.
+
+When the plan is purely analytical (e.g. a code review or feasibility study
+with no implementation), skip the **Approach** and **Task List** sections and
+lead with **Findings** instead, keeping the rest of the structure.
+
+## Common Rationalizations
+
+| Rationalization | Reality |
+|---|---|
+| "I'll figure it out as I go" | That's how you end up with a tangled mess and rework. 10 minutes of planning saves hours. |
+| "The tasks are obvious" | Write them down anyway. Explicit tasks surface hidden dependencies and forgotten edge cases. |
+| "Planning is overhead" | Planning is the task. Implementation without a plan is just typing. |
+| "I can hold it all in my head" | Context windows are finite. Written plans survive session boundaries and compaction. |
+
+## Verification Checklist
+
+Before delivering the plan, confirm:
+
+- [ ] Every task has acceptance criteria
+- [ ] Every task has a verification step
+- [ ] Task dependencies are identified and ordered correctly
+- [ ] No task is XL or larger — break it down instead
+- [ ] Checkpoints exist after every 2-3 tasks
+- [ ] The response states the plan's path (saved or suggested) and suggests
+      `/review-plan` and `/implement-plan`
+- [ ] The plan is ready for human review
@@ -0,0 +1,47 @@
+---
+name: resolve-git-conflicts
+description: Conflict resolution flow — understand the local git conflicts, present a resolution plan, and resolve them after the user approves it. Never continues the rebase. Use it when the repo has unresolved conflicts (rebase, merge, cherry-pick) or the user asks to resolve them.
+---
+
+# Resolve Git Conflicts
+
+Resolve conflicts in the local repository. The user handles finishing the
+rebase themselves — you must **never** run `git rebase --continue`,
+`git rebase --skip`, `git merge --continue`, or anything similar.
+
+## When to use
+
+- The repository has unresolved conflicts — during a rebase, merge, or
+  cherry-pick — whether the user asks about them or not.
+- The user asks to resolve conflicts, in any phrasing: "fix the merge
+  conflicts", "resolve these", "what's conflicting here?".
+
+## Phase 1 — Understand the problem (read-only)
+
+1. Run `git status` to detect the conflict state (rebase, merge, cherry-pick, etc.) and list conflicted files.
+2. For each conflicted (unmerged) file, understand the situation **without modifying anything**:
+   - Read the file and identify the conflict markers (`<<<<<<<`, `=======`, `>>>>>>>`).
+   - Inspect both sides — `git show <ours>:<file>` and `git show <theirs>:<file>` — plus `git log`/`git show` on the commits involved to understand intent.
+   - Identify what each side changed and why, and how they should be combined.
+
+## Phase 2 — Present the resolution plan
+
+3. **Present a clear plan to the user before touching any file.** For each conflicted file, state:
+   - What each side changed and why.
+   - Your proposed resolution and the reasoning behind it.
+   - How the two sides are combined (both additive → merge; both modify the same code → keep the semantically correct version, merging intent from both sides when clear from code and context).
+4. **Ask the user only when genuinely unclear.** Do not ask about anything you can determine yourself from the code, commit messages, or context. Only decisions that are not determinable and change the outcome (e.g. conflicting product decisions, which side to discard) warrant a question. **Collect all such questions together in an "Open Questions" section at the end of the plan**, so the user has full context to answer them properly.
+5. **Wait for the user to accept the plan** (and answer any open questions) before editing, staging, or otherwise modifying anything.
+
+## Phase 3 — Execute
+
+6. Resolve each conflicted file by editing the file to the agreed merged content and removing all conflict markers.
+
+## Phase 4 — Stage and verify
+
+7. **Stage every resolved file** with `git add <file>`. Do not stage unrelated untracked files unless clearly part of the resolution.
+8. Verify no conflict markers remain (search for `<<<<<<<` / `>>>>>>>` in resolved files) and that `git status` shows no unmerged paths.
+
+## Phase 5 — Report
+
+9. Briefly report the conflict state, how each conflicted file was resolved (and any answers received to open questions), and stop — do **not** run `git rebase --continue` or any other continuation command.
@@ -0,0 +1,73 @@
+---
+name: review-code
+description: Code review flow — review a diff, PR, or code change, delegating the review to a subagent that follows the code-review-criteria skill. Use it when the user asks to review code or a PR, in any phrasing.
+---
+
+# Review Code
+
+Act as a senior software engineer and perform a thorough code review.
+
+## When to use
+
+- The user asks to review code, in any phrasing: "review this diff",
+  "review the PR", "check my changes", "code review" — or runs
+  `/review-code`.
+- A commit, branch, PR, or diff is ready and the user wants it assessed
+  before merge.
+
+## Instructions
+
+1. **Determine what is being reviewed** from the user context: a working-tree
+   diff, a commit range, a branch, a PR (number or URL), or specific files. If
+   the target is ambiguous, ask before reviewing.
+2. Delegate the review to the `general` subagent (via the task tool), unless the
+   user specifies another agent. Include in the prompt the
+   **`code-review-criteria`** skill name and all user context.
+3. When the subagent returns, output the review to the user verbatim. Do not
+   summarize it and do not act on its findings.
+4. Right after the review, suggest how to proceed based on the findings. These
+   are suggestions — the user decides:
+   - **Approve (no required changes):** say so — there is nothing to address.
+   - **Minor findings (nits):** applying them directly as-is is fine once the
+     review is done — no plan needed.
+   - **Substantive findings:** suggest `/make-a-plan` to make a plan to address
+     them.
+
+### Hard rule — read-only while reviewing
+
+This flow is read-only **for the duration of the review**: from the moment it
+starts until the user considers the review finished (including any feedback,
+questions, or clarifications about it). During that period, never fix,
+implement, edit files or create commits — not even "obvious" fixes derived from
+the findings. Once the user explicitly states the review is done (or moves on to
+a different task), this rule no longer applies and you act as a normal build
+agent again.
+
+## Instructions for the subagent
+
+1. Load the **`code-review-criteria`** skill and follow its process and output
+   format.
+2. Read `AGENTS.md` (if present) and follow its instructions for finding and
+   reading all related testing documentation from memories before reviewing.
+3. Return in your final message the COMPLETE review, verbatim, exactly as the
+   skill instructs it to be produced. Do not summarize it — include the full
+   structured review.
+
+### Strong rules for the subagent
+
+1. Do not invent problems. Every finding must be real and actionable.
+2. Read-only: do not modify any file and do not create a commit — reviewing
+   never writes.
+3. Be specific and constructive. "This could be better" is not helpful — explain
+   why and how.
+4. Prioritize by impact. One structural issue outweighs ten nits.
+5. Missing tests are an issue, not a suggestion. Report as a severity-tagged
+   finding — never as a recommendation.
+6. Skip generated files, lockfile-only changes, and unrelated modifications
+   unless they introduce security risks.
+
+## User context
+
+Extra context in the user's invocation (the message that triggered this skill)
+plays the role command arguments play elsewhere: for example, a PR number or
+URL, a commit range, specific files, or a different agent to run the review.
@@ -0,0 +1,71 @@
+---
+name: review-plan
+description: Plan review flow — evaluate an implementation plan before it is executed, delegating the review to a subagent that follows the plan-review-criteria skill. Use it when the user asks to review a plan, in any phrasing.
+---
+
+# Review Plan
+
+Act as a senior software engineer and perform a thorough review of an
+implementation plan.
+
+## When to use
+
+- The user asks to review a plan, in any phrasing: "review this plan",
+  "does this plan look right?", "second opinion on the plan" — or runs
+  `/review-plan`.
+- A plan was just produced (typically by `/make-a-plan`) and the user
+  wants it evaluated before executing it.
+
+## Instructions
+
+1. **Determine the plan under review** from the session context (for example, a
+   plan just produced by `/make-a-plan`) or from a plan file path given by the
+   user (typically under `.agents/plans/`). If a file path is given, read the
+   file first so the complete plan is in context.
+2. Delegate the review to the `general` subagent (via the task tool), unless the
+   user specifies another agent. Include in the prompt the
+   **`plan-review-criteria`** skill name and all user context.
+3. When the subagent returns, output the review to the user verbatim. Do not
+   summarize it and do not act on its findings.
+4. Right after the review, suggest the next step based on the verdict. These
+   are suggestions — the user decides, and any instruction overrides them:
+   - **Approve** → suggest `/implement-plan` to execute it.
+   - **Request changes** → suggest `/make-a-plan` to make a plan to address the
+     findings.
+
+### Hard rule — read-only while reviewing
+
+This flow is read-only **for the duration of the review**: from the moment it
+starts until the user considers the review finished (including any feedback,
+questions, or clarifications about it). During that period, never fix,
+implement, edit files or create commits — not even "obvious" fixes derived from
+the findings. Once the user explicitly states the review is done (or moves on to
+a different task), this rule no longer applies and you act as a normal build
+agent again.
+
+## Instructions for the subagent
+
+1. Load the **`plan-review-criteria`** skill and follow its process and output
+   format.
+2. Read `AGENTS.md` (if present) and follow its instructions for finding and
+   reading all related documentation and testing memories before reviewing.
+3. Return in your final message the COMPLETE review, verbatim, exactly as the
+   skill instructs it to be produced. Do not summarize it — include the full
+   structured review.
+
+### Strong rules for the subagent
+
+1. Do not invent problems. Every finding must be real and actionable.
+2. Read-only: do not modify any file and do not create a commit — reviewing
+   never writes.
+3. Be specific and constructive. "This could be better" is not helpful — explain
+   why and how.
+4. Prioritize by impact. One structural issue outweighs ten nits.
+5. Judge the plan as the implementer would: every task executable without
+   guessing, ordering follows the dependency graph, risks named.
+
+## User context
+
+Extra context in the user's invocation (the message that triggered this skill)
+plays the role command arguments play elsewhere: for example, a plan file path
+to review, or a different agent to run the review.
@@ -9,6 +9,11 @@ metadata: {"clawdbot":{"emoji":"🔎","requires":{"bins":["rg"]},"install":[{"id

 Fast, smart recursive search. Respects `.gitignore` by default.

+## When to use
+
+- Searching file contents across the repo for regex patterns — the
+  default code search, respects `.gitignore`.
+
 ## Quick Start

 ### Basic search
@@ -9,6 +9,13 @@ Apply the ASD-STE100 standard to all prose you produce in this task. Do not anno

 Compliance note (for you, not for output): the official specification and its dictionary are copyright ASD. This skill encodes paraphrased rules and a publicly sourced word list. For certified aerospace/defense deliverables, tell the user that full compliance requires the free official specification (asd-ste100.org) and a human sign-off. Never claim certified compliance.

+## When to use
+
+Only when the user explicitly invokes it: they type `/ste`, or say "use
+the ste skill" / "apply ASD-STE100". Requests like "simplify this",
+"make it clearer", or "shorter sentences" do NOT invoke it — respond
+normally unless it is named.
+
 ## Step 0 — Classify the text

 Before writing a single sentence, decide: is this **procedural** text (instructions someone follows) or **descriptive** text (explanation, background, description)? Every limit below depends on this. Mixed documents get classified section by section.
@@ -11,6 +11,12 @@ Fetch information from Taiga public API for the **Penpot** project

 **No authentication required** — only public project data is accessed.

+## When to use
+
+- The user asks about Penpot issues, user stories, or tasks tracked in
+  Taiga — fetch them via the public API (project id 345963), no
+  authentication needed.
+
 ## Prerequisites

 - `python3` — the `scripts/taiga.py` CLI script is self-contained (stdlib only)
@@ -357,6 +357,39 @@ Insert the new version section right after the `# CHANGELOG` header (before
 the previous version entry). Use the `edit` tool with enough context to make
 a unique match.

+### 8b. Propose and populate the `:rocket: Epics and highlights` subsection
+
+After inserting the version section, proactively create or populate the
+`### :rocket: Epics and highlights` subsection. This section surfaces the
+most impactful changes for self-hosted users checking for updates.
+
+**When to create:** If the version section does not already have a
+`### :rocket: Epics and highlights` subsection, create one. Place it before
+`### :sparkles:` (matching existing order in CHANGES.md).
+
+**How to identify highlights:** Review the `:sparkles:` entries for the
+version and select 2–5 of the most impactful/user-visible ones. Criteria:
+- New user-visible features (not internal refactors)
+- Significant capability additions
+- Items that create "FOMO" for self-hosted users on older versions
+
+**Use release notes as hints:** Check
+`frontend/src/app/main/ui/releases/v2_<MINOR>.cljs` for the corresponding
+version. The slide titles and feature descriptions there are curated
+marketing content indicating what the team considers highlight-worthy. Match
+those themes to changelog entries. Treat these files as optional hints — they
+may not exist for every version.
+
+**Format requirement:** Every `:rocket:` entry MUST follow the standard
+changelog format with issue/PR references:
+```
+- <description> [#<ISSUE>](https://github.com/penpot/penpot/issues/<ISSUE>) (PR: [#<PR>](https://github.com/penpot/penpot/pull/<PR>))
+```
+An entry without issue AND PR references is a highlight gap (warning, not an anomaly).
+
+**Preserve existing entries:** If the `:rocket:` section already exists from
+a prior run, preserve its entries. Do not remove or rewrite them.
+
 ### 9. Verify

 Read the top of `CHANGES.md` and confirm:
@@ -468,9 +501,8 @@ Markdown viewer.
 ## What is an anomaly

 **An anomaly is a milestone-mismatch between an issue and its referenced
-PR.** It indicates that the changelog claim "this issue is fixed by this PR,
-all in milestone M" is inconsistent with the actual milestone assignments.
-There are exactly two types:
+PR.** There are two anomaly types, plus two highlight gaps (warnings that
+do not count toward the anomaly total):

 1. **Issue is in the milestone, but its referenced PR is in a different
   milestone (or has no milestone).** The changelog claims a fix in this
@@ -486,6 +518,13 @@ There are exactly two types:
   PR that closes an issue with no milestone references an issue from
   another (probably private) project; that is expected and the issue is
   not part of this changelog. Do not report it.
+3. **missing-highlights (gap):** A released X.Y.0 version section has no
+   `### :rocket: Epics and highlights` subsection. Patches (X.Y.Z) never
+   carry highlights, so only minors/majors are checked.
+4. **missing-highlight-reference (gap):** A `:rocket:` entry lacks the
+   required issue AND PR references. Every highlight entry must follow the
+   standard changelog format with `[#ISSUE]` and `(PR: [#PR])` links
+   (multi-PR `(PR: [#A](...), [#B](...))` accepted).

 **Anything else is not an anomaly.** Other discrepancies (exclusion
 labels on in-changelog issues, missing valid issues, unmerged PR
@@ -653,6 +692,40 @@ for pr_num in sorted(changelog_prs):
                'issue_milestone': issue_ms,  # may be None
            })

+# --- Type C: released X.Y.0 version sections without :rocket: subsection ---
+# Patches (X.Y.Z with Z != 0) never carry :rocket: by design — only minors/majors (X.Y.0).
+anomalies_c = []  # list of version strings
+rocket_heading_re = re.compile(r'^### :rocket:', re.MULTILINE)
+version_sections = re.split(r'(?=^## \d+\.\d+\.\d+)', content, flags=re.MULTILINE)
+for vs in version_sections:
+    m = re.match(r'^## (\d+\.\d+\.\d+)(.*)', vs)
+    if not m: continue
+    ver, suffix = m.group(1), m.group(2)
+    if 'unreleased' in suffix.lower(): continue
+    if ver.split('.')[2] != '0': continue
+    if not rocket_heading_re.search(vs):
+        anomalies_c.append(ver)
+
+# --- Type D: :rocket: entries without issue AND PR references ---
+# Both are required: `[#ISSUE](.../issues/N)` and `(PR: [#PR](.../pull/M))`.
+# Multi-PR entries `(PR: [#A](...), [#B](...))` are accepted.
+anomalies_d = []  # list of dicts: {version, line}
+issue_ref_re = re.compile(r'\[#\d+\]\(https://github\.com/penpot/penpot/issues/\d+\)')
+pr_ref_re = re.compile(r'\(PR:\s*\[#\d+\]\(https://github\.com/penpot/penpot/pull/\d+\)(\s*,\s*\[#\d+\]\(https://github\.com/penpot/penpot/pull/\d+\))*\)')
+for vs in version_sections:
+    m = re.match(r'^## (\d+\.\d+\.\d+)(.*)', vs)
+    if not m: continue
+    ver = m.group(1)
+    rocket_match = rocket_heading_re.search(vs)
+    if not rocket_match: continue
+    # Extract the :rocket: subsection body (up to next ### or ##)
+    rocket_body = vs[rocket_match.end():]
+    rocket_body = re.split(r'(?m)^#{2,3}\s', rocket_body)[0]
+    for line in rocket_body.splitlines():
+        line = line.strip()
+        if line.startswith('- ') and not (issue_ref_re.search(line) and pr_ref_re.search(line)):
+            anomalies_d.append({'version': ver, 'line': line[:100]})
+
 # --- Write report ---
 def fmt_ms(ms):
    return ms if ms else "_none_"
@@ -664,13 +737,17 @@ with open(OUTPUT, 'w') as f:

    n_a = len(anomalies_a)
    n_b = len(anomalies_b)
+    n_c = len(anomalies_c)
+    n_d = len(anomalies_d)

    f.write('## Summary\n\n')
    f.write(f'- **Issue in {MILESTONE}, referenced PR in different milestone or no milestone:** {n_a}\n')
    f.write(f'- **PR in {MILESTONE}, closing issue in a different milestone:** {n_b}\n')
-    f.write(f'- **Total anomalies:** {n_a + n_b}\n\n')
+    f.write(f'- **Total anomalies:** {n_a + n_b}\n')
+    f.write(f'- **Released X.Y.0 version missing :rocket: section (gap):** {n_c}\n')
+    f.write(f'- **:rocket: entry without issue AND PR references (gap):** {n_d}\n\n')

-    # --- Anomalies section ---
+    # --- Anomalies section (milestone mismatches only) ---
    if n_a or n_b:
        f.write('## Anomalies\n\n')
        f.write('These are milestone mismatches between an issue in the changelog '
@@ -709,9 +786,37 @@ with open(OUTPUT, 'w') as f:
                    badge = '🔴' if e['issue_milestone'] is None else '⚠️'
                    f.write(f'  - {badge} Closing {issue_link(e["issue"])} is in milestone **{ms_label}** (expected: {MILESTONE})\n')
                f.write('\n')
+
    else:
        f.write('✅ No anomalies found. All (issue, PR) pairs in the changelog have aligned milestone assignments.\n\n')

+    # --- Highlight gaps (warnings, not anomalies) ---
+    if n_c or n_d:
+        f.write('## Highlight gaps\n\n')
+        f.write('These are warnings, not anomalies: they do not affect the '
+                'milestone-mismatch total above. They track `:rocket:` coverage '
+                'across all released X.Y.0 versions. Historical entries (e.g. '
+                'Taiga links) predate the current reference convention and are '
+                'expected to appear here.\n\n')
+
+        if n_c:
+            f.write(f'### Released X.Y.0 version missing :rocket: section\n\n')
+            f.write('These released minors/majors have no `### :rocket: Epics and highlights` subsection. '
+                    'Add highlights to help self-hosted users understand what they are missing.\n\n')
+            for ver in anomalies_c:
+                f.write(f'- Version **{ver}**\n')
+            f.write('\n')
+
+        if n_d:
+            f.write(f'### :rocket: entry without issue AND PR references\n\n')
+            f.write('These highlight entries lack the required issue AND PR references. '
+                    'Add `[#ISSUE](...)` and `(PR: [#PR](...))` links.\n\n')
+            for d in anomalies_d:
+                f.write(f'- **{d["version"]}**: `{d["line"]}`\n')
+            f.write('\n')
+    elif not (n_a or n_b):
+        f.write('✅ No highlight gaps found. All released X.Y.0 versions have properly referenced :rocket: entries.\n\n')
+
    # --- Context ---
    f.write('---\n\n')
    f.write('## Context\n\n')
@@ -726,8 +831,7 @@ print(f"Anomaly report written to {OUTPUT}")
 PYEOF
 ```

-This generates `CHANGES-ISSUES.md` containing **only the anomalies** —
-milestone mismatches between issues and their referenced PRs:
+This generates `CHANGES-ISSUES.md` containing anomalies and highlight gaps:

 1. **Issue in milestone, referenced PR in different milestone or no milestone** —
   the changelog claims a fix here, but the PR is released elsewhere.
@@ -736,6 +840,13 @@ milestone mismatches between issues and their referenced PRs:
   (An issue with *no* milestone belongs to another, probably private,
   project — milestones are only required on the "Main" project — so it is
   neither an anomaly nor a changelog candidate.)
+3. **missing-highlights (gap, warning)** — a released X.Y.0 version section
+   has no `### :rocket: Epics and highlights` subsection. Patches (X.Y.Z)
+   never carry highlights.
+4. **missing-highlight-reference (gap, warning)** — a `:rocket:` entry lacks
+   the required issue AND PR references.
+
+Gaps do not count toward the anomaly total.

 **Rule violations are not in the report** — they are workflow errors the
 LLM must fix directly in `CHANGES.md` during step 6a (pre-flight checks).
@@ -809,10 +920,13 @@ self-contained and clickable in any Markdown viewer.
  issue from a different project or context. If the PR title and issue title
  are clearly unrelated, or the PR predates the issue by years, treat it as a
  data glitch and skip it.
- **Anomaly = milestone mismatch only.** The report contains only milestone
-  mismatches: (1) the issue is in this milestone but the referenced PR is
-  in a different milestone (or unassigned), and (2) the PR is in this
-  milestone but the issue it closes is in a different milestone. An
+- **Anomaly = milestone mismatch only; gaps are warnings.** The report's
+  anomaly total counts only milestone mismatches: (1) the issue is in this
+  milestone but the referenced PR is in a different milestone (or unassigned),
+  and (2) the PR is in this milestone but the issue it closes is in a
+  different milestone. `:rocket:` highlight gaps (missing section on a
+  released X.Y.0, entry without issue AND PR references) are reported in a
+  separate `Highlight gaps` section and never count toward the anomaly total. An
  *unassigned* (milestone-less) issue closed by a milestone PR is **not**
  an anomaly: milestones are required only for the "Main" project, so such
  issues come from another (probably private) project and are not changelog
@@ -0,0 +1 @@
+../.agents/skills
@@ -0,0 +1,44 @@
+name: _ADHOC
+
+run-name: >-
+  _ADHOC (${{ inputs.gh_ref }}${{ inputs.nitrate_ref != '' && format(' / nitrate:{0}', inputs.nitrate_ref) || '' }})
+
+on:
+  workflow_dispatch:
+    inputs:
+      gh_ref:
+        description: 'Branch/ref to build in penpot/penpot'
+        type: string
+        required: true
+      nitrate_ref:
+        description: 'Branch/ref to build admin-console in penpot/penpot-nitrate (defaults to gh_ref)'
+        type: string
+        required: false
+      force:
+        description: 'Rebuild and overwrite even if already built/promoted'
+        type: boolean
+        required: false
+        default: false
+
+jobs:
+  build-bundle:
+    uses: ./.github/workflows/build-bundle.yml
+    secrets: inherit
+    with:
+      gh_ref: ${{ inputs.gh_ref }}
+      force: ${{ inputs.force }}
+
+  build-docker:
+    needs: build-bundle
+    uses: ./.github/workflows/build-docker.yml
+    secrets: inherit
+    with:
+      gh_ref: ${{ inputs.gh_ref }}
+      force: ${{ inputs.force }}
+
+  build-docker-admin-console:
+    uses: ./.github/workflows/build-docker-admin-console.yml
+    secrets: inherit
+    with:
+      gh_ref: ${{ inputs.nitrate_ref || inputs.gh_ref }}
+      force: ${{ inputs.force }}
@@ -9,6 +9,11 @@ on:
        type: string
        required: true
        default: 'develop'
+      force:
+        description: 'Rebuild and overwrite even if this version already exists in S3'
+        type: boolean
+        required: false
+        default: false
  workflow_call:
    inputs:
      gh_ref:
@@ -16,6 +21,11 @@ on:
        type: string
        required: true
        default: 'develop'
+      force:
+        description: 'Rebuild and overwrite even if this version already exists in S3'
+        type: boolean
+        required: false
+        default: false

 # Literal group name: under `workflow_call`, `github.workflow` resolves to the
 # caller's workflow, which put this workflow and the other reusable one called
@@ -34,6 +44,8 @@ jobs:
    outputs:
      gh_ref: ${{ steps.vars.outputs.gh_ref }}
      bundle_version: ${{ steps.vars.outputs.bundle_version }}
+      sha: ${{ steps.vars.outputs.sha }}
+      commit_title: ${{ steps.vars.outputs.commit_title }}
      exists: ${{ steps.check.outputs.exists }}

    steps:
@@ -48,10 +60,12 @@ jobs:
        run: |
          echo "gh_ref=${{ inputs.gh_ref || github.ref_name }}" >> $GITHUB_OUTPUT
          echo "bundle_version=$(git describe --tags --always)" >> $GITHUB_OUTPUT
+          echo "sha=$(git rev-parse --short=12 HEAD)" >> $GITHUB_OUTPUT
+          echo "commit_title=$(git log -1 --pretty=%s)" >> $GITHUB_OUTPUT

      # The uploaded zip carries its version as S3 metadata. If the
      # existing object was already built from this same commit, the
-      # whole build job is skipped.
+      # whole build job is skipped. `force` bypasses this check entirely.
      - name: Check if this bundle is already built
        id: check
        env:
@@ -59,6 +73,16 @@ jobs:
          AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
          AWS_DEFAULT_REGION: ${{ secrets.AWS_REGION }}
        run: |
+          if [ "${{ inputs.force }}" = "true" ]; then
+            echo "exists=false" >> $GITHUB_OUTPUT
+            {
+              echo "### 🔁 Bundle build forced"
+              echo ""
+              echo "\`force: true\` — skipping the S3 version check."
+            } >> "$GITHUB_STEP_SUMMARY"
+            exit 0
+          fi
+
          EXISTING_VERSION=$(aws s3api head-object \
            --bucket ${{ secrets.S3_BUCKET }} \
            --key "penpot-${{ steps.vars.outputs.gh_ref }}.zip" \
@@ -117,6 +141,16 @@ jobs:
            s3://${{ secrets.S3_BUCKET }}/penpot-${{ needs.check.outputs.gh_ref }}.zip \
            --metadata bundle-version=${{ needs.check.outputs.bundle_version }}

+      - name: Write step summary
+        run: |
+          {
+            echo "### ✅ Bundle built"
+            echo ""
+            echo "- Version: \`${{ needs.check.outputs.bundle_version }}\` (\`git describe --tags --always\`)"
+            echo "- Commit: [\`${{ needs.check.outputs.sha }}\`](https://github.com/${{ github.repository }}/commit/${{ needs.check.outputs.sha }}) — ${{ needs.check.outputs.commit_title }}"
+            echo "- Built at: $(date -u +'%Y-%m-%d %H:%M:%S UTC')"
+          } >> "$GITHUB_STEP_SUMMARY"
+
  # ── 3. Single failure notification for the whole workflow ─────────────
  notify:
    name: Notify failure
@@ -1,7 +1,16 @@
 name: _DEVELOP

+run-name: >-
+  _DEVELOP (develop @ ${{ github.sha }})
+
 on:
  workflow_dispatch:
+    inputs:
+      force:
+        description: 'Rebuild and overwrite even if already built/promoted'
+        type: boolean
+        required: false
+        default: false
  schedule:
    - cron: '16 5-20 * * 1-5'

@@ -15,6 +24,7 @@ jobs:
    secrets: inherit
    with:
      gh_ref: "develop"
+      force: ${{ inputs.force || false }}

  build-docker:
    needs: build-bundle
@@ -22,9 +32,11 @@ jobs:
    secrets: inherit
    with:
      gh_ref: "develop"
+      force: ${{ inputs.force || false }}

  build-docker-admin-console:
    uses: ./.github/workflows/build-docker-admin-console.yml
    secrets: inherit
    with:
      gh_ref: "develop"
+      force: ${{ inputs.force || false }}
@@ -13,6 +13,11 @@ on:
        type: string
        required: false
        default: 'develop'
+      force:
+        description: 'Rebuild and overwrite even if already built'
+        type: boolean
+        required: false
+        default: false
  workflow_call:
    inputs:
      gh_ref:
@@ -24,6 +29,11 @@ on:
        type: string
        required: false
        default: 'develop'
+      force:
+        description: 'Rebuild and overwrite even if already built'
+        type: boolean
+        required: false
+        default: false
    secrets:
      ORG_WORKFLOW_TOKEN:
        description: 'Token with Actions write access on penpot-nitrate'
@@ -47,6 +57,7 @@ jobs:

          gh workflow run "$WORKFLOW" --repo "$REPO" --ref "$DISPATCH_REF" \
            -f gh_ref="$GH_REF" \
+            -f force="${{ inputs.force }}" \
            -f caller_run_id="$DISTINCT_ID" \
            -f caller_run_url="$CALLER_URL"

@@ -6,7 +6,7 @@ on:
 jobs:
  build-and-push:
    name: Build and push DevEnv Docker image
-    runs-on: penpot-extended-runner
+    runs-on: penpot-standar-runner

    steps:
      - name: Set common environment variables
@@ -8,6 +8,11 @@ on:
        type: string
        required: true
        default: 'develop'
+      force:
+        description: 'Rebuild and overwrite even if this sha is already promoted'
+        type: boolean
+        required: false
+        default: false
  workflow_call:
    inputs:
      gh_ref:
@@ -15,6 +20,11 @@ on:
        type: string
        required: true
        default: 'develop'
+      force:
+        description: 'Rebuild and overwrite even if this sha is already promoted'
+        type: boolean
+        required: false
+        default: false

 # Literal group name: under `workflow_call`, `github.workflow` resolves to the
 # caller's workflow, which put this workflow and the other reusable one called
@@ -36,12 +46,13 @@ jobs:
  # ── 1. Resolve the build key and check the whole set at once ───────────
  prepare:
    name: Prepare
-    runs-on: penpot-extended-runner
+    runs-on: penpot-standar-runner
    timeout-minutes: 15
    outputs:
      gh_ref: ${{ steps.vars.outputs.gh_ref }}
      bundle_version: ${{ steps.vars.outputs.bundle_version }}
      sha: ${{ steps.vars.outputs.sha }}
+      commit_title: ${{ steps.vars.outputs.commit_title }}
      exists: ${{ steps.check.outputs.exists }}

    steps:
@@ -60,6 +71,7 @@ jobs:
          GH_REF="${{ inputs.gh_ref || github.ref_name }}"
          echo "gh_ref=$GH_REF" >> $GITHUB_OUTPUT
          echo "sha=$(git rev-parse --short=12 HEAD)" >> $GITHUB_OUTPUT
+          echo "commit_title=$(git log -1 --pretty=%s)" >> $GITHUB_OUTPUT

          BUNDLE_VERSION=$(aws s3api head-object \
            --bucket ${{ secrets.S3_BUCKET }} \
@@ -71,7 +83,8 @@ jobs:
      # The image set is a single block, so a single set-level check is
      # enough: `promote` drops a marker object in S3 only after every
      # image was built AND every branch tag was moved. Marker present
-      # means there is nothing at all to do for this commit.
+      # means there is nothing at all to do for this commit. `force`
+      # bypasses this check entirely.
      - name: Check if this image set is already built
        id: check
        env:
@@ -79,6 +92,21 @@ jobs:
          AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
          AWS_DEFAULT_REGION: ${{ secrets.AWS_REGION }}
        run: |
+          if [ "${{ inputs.force }}" = "true" ]; then
+            echo "exists=false" >> $GITHUB_OUTPUT
+            mkdir -p "$BUNDLE_CACHE"
+            find "$BUNDLE_CACHE" -type f -mtime +1 -delete || true
+            ZIP="$BUNDLE_CACHE/penpot-${{ steps.vars.outputs.bundle_version }}.zip"
+            aws s3 cp "s3://${{ secrets.S3_BUCKET }}/penpot-${{ steps.vars.outputs.gh_ref }}.zip" "$ZIP.$$.tmp"
+            mv "$ZIP.$$.tmp" "$ZIP"
+            {
+              echo "### 🔁 Image set build forced"
+              echo ""
+              echo "\`force: true\` — skipping the S3 marker check."
+            } >> "$GITHUB_STEP_SUMMARY"
+            exit 0
+          fi
+
          if aws s3api head-object \
               --bucket ${{ secrets.S3_BUCKET }} \
               --key "markers/images-sha-${{ steps.vars.outputs.sha }}" \
@@ -107,7 +135,7 @@ jobs:
  # ── 2. One build per image, in parallel, only when needed ──────────────
  build:
    name: Build ${{ matrix.image }}
-    runs-on: penpot-extended-runner
+    runs-on: penpot-standar-runner
    timeout-minutes: 60
    needs: prepare
    if: needs.prepare.outputs.exists == 'false'
@@ -138,7 +166,7 @@ jobs:
          username: ${{ secrets.DOCKER_USERNAME }}
          password: ${{ secrets.DOCKER_PASSWORD }}

-      # To avoid the “429 Too Many Requests” error when downloading
+      # To avoid the "429 Too Many Requests" error when downloading
      # images from DockerHub for unregistered users.
      # https://docs.docker.com/docker-hub/usage/
      - name: Login to DockerHub Registry
@@ -220,7 +248,7 @@ jobs:
  # the S3 marker guarantees the branch tags were already moved.
  promote:
    name: Promote image set
-    runs-on: penpot-extended-runner
+    runs-on: penpot-standar-runner
    timeout-minutes: 10
    needs: [prepare, build]

@@ -258,16 +286,23 @@ jobs:
        run: |
          echo "${{ github.run_id }}" | aws s3 cp - \
            "s3://${{ secrets.S3_BUCKET }}/markers/images-sha-${{ needs.prepare.outputs.sha }}"
+
+      - name: Write step summary
+        run: |
          {
            echo "### ✅ Image set promoted"
            echo ""
+            echo "- Version: \`${{ needs.prepare.outputs.bundle_version }}\` (\`git describe --tags --always\`)"
+            echo "- Commit: [\`${{ needs.prepare.outputs.sha }}\`](https://github.com/${{ github.repository }}/commit/${{ needs.prepare.outputs.sha }}) — ${{ needs.prepare.outputs.commit_title }}"
+            echo "- Built at: $(date -u +'%Y-%m-%d %H:%M:%S UTC')"
+            echo ""
            echo "All \`:${{ needs.prepare.outputs.gh_ref }}\` tags now point to \`sha-${{ needs.prepare.outputs.sha }}\`."
          } >> "$GITHUB_STEP_SUMMARY"

  # ── 4. Single failure notification for the whole workflow ─────────────
  notify:
    name: Notify failure
-    runs-on: penpot-extended-runner
+    runs-on: penpot-standar-runner
    timeout-minutes: 5
    needs: [prepare, build, promote]
    if: failure()
@@ -1,7 +1,16 @@
 name: _STAGING

+run-name: >-
+  _STAGING (staging)
+
 on:
  workflow_dispatch:
+    inputs:
+      force:
+        description: 'Rebuild and overwrite even if already built/promoted'
+        type: boolean
+        required: false
+        default: false
  schedule:
    - cron: '36 5-20 * * 1-5'

@@ -15,6 +24,7 @@ jobs:
    secrets: inherit
    with:
      gh_ref: "staging"
+      force: ${{ inputs.force || false }}

  build-docker:
    needs: build-bundle
@@ -22,9 +32,11 @@ jobs:
    secrets: inherit
    with:
      gh_ref: "staging"
+      force: ${{ inputs.force || false }}

  build-docker-admin-console:
    uses: ./.github/workflows/build-docker-admin-console.yml
    secrets: inherit
    with:
      gh_ref: "staging"
+      force: ${{ inputs.force || false }}
@@ -1,7 +1,16 @@
 name: _TAG

+run-name: >-
+  _TAG (${{ github.ref_name }} @ ${{ github.sha }})
+
 on:
  workflow_dispatch:
+    inputs:
+      force:
+        description: 'Rebuild and overwrite even if already built/promoted (manual re-releases only)'
+        type: boolean
+        required: false
+        default: false
  push:
    tags:
      - '*'
@@ -18,6 +27,7 @@ jobs:
    secrets: inherit
    with:
      gh_ref: ${{ github.ref_name }}
+      force: ${{ inputs.force || false }}

  build-docker:
    needs: build-bundle
@@ -25,16 +35,18 @@ jobs:
    secrets: inherit
    with:
      gh_ref: ${{ github.ref_name }}
+      force: ${{ inputs.force || false }}

  build-docker-admin-console:
    uses: ./.github/workflows/build-docker-admin-console.yml
    secrets: inherit
    with:
      gh_ref: ${{ github.ref_name }}
+      force: ${{ inputs.force || false }}

  notify:
    name: Notifications
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-latest
    needs:
      - build-docker
      - build-docker-admin-console
@@ -19,7 +19,7 @@ permissions:

 jobs:
  release:
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-latest
    outputs:
      version: ${{ steps.vars.outputs.gh_ref }}
      release_notes: ${{ steps.extract_release_notes.outputs.release_notes }}
@@ -32,7 +32,7 @@ jobs:
  test-exporter:
    if: ${{ !github.event.pull_request.draft }}
    name: "Exporter Tests"
-    runs-on: penpot-runner-02
+    runs-on: penpot-extended-runner
    container:
      image: penpotapp/devenv:latest
      volumes:
@@ -24,7 +24,6 @@ opencode.json
 !AGENTS.md
 !CODE_OF_CONDUCT.md
 !SECURITY.md
-!HIGHLIGHTS.md
 /*.png
 /*.svg
 /*.sql
@@ -103,7 +102,7 @@ opencode.json
 /.playwright-mcp
 /.devenv/mcp/
 /opencode.json
-/.opencode/plans
+/.agents/plans
 /.opencode/reports
 /.opencode/prompts
 /.ci-logs
@@ -0,0 +1,10 @@
+---
+description: Create a PR for the current task branch or update an existing one — loads and follows the create-pr skill
+agent: build
+---
+
+Load the **`create-pr`** skill and follow it as your only instruction.
+
+## User input, overrides and additional context
+
+$ARGUMENTS
@@ -1,39 +1,10 @@
 ---
-description: Execute a ready plan end-to-end — create a GitHub issue, branch issue-NNNN, implement the plan, then commit via the create-commit skill
+description: Execute a ready plan — task checklist, your confirmation, then all tasks with one commit (default) or step by step with a commit and a pause per task; creates issue + branch when on a base branch, or commits on the current branch with "direct" — loads and follows the implement-plan skill
 agent: build
 ---

-This command is run once a plan is ready (for example, from plan mode). Execute
-the plan already prepared in the current session context. Follow these steps in order.
+Load the **`implement-plan`** skill and follow it as your only instruction.

-## 1. Create the issue
+## User input, overrides and additional context

-Use the **`create-issue`** skill, following the *Creating Issues from Draft Body*
-flow in `mem:workflow/creating-issues`. Derive the issue title and body from the
-plan. Capture the new issue's number — call it **NNNN** (needed for the branch
-name and the commit reference).
-
-## 2. Create the branch
-
-Create and switch to a branch named after the issue:
-
-```
-git checkout -b issue-NNNN
-```
-
-(Replace NNNN with the issue number from step 1.)
-
-## 3. Execute the plan
-
-Implement the prepared plan from the session context. Work methodically, keeping
-changes focused on what the issue requires. Do not commit — the commit happens in
-step 4.
-
-## 4. Commit with the create-commit skill
-
-After the implementation is complete, load the **`create-commit`** skill and
-follow its workflow to commit the changes. Provide a brief summary of what was
-implemented and why, the issue reference (`issue-NNNN`), and the model name you
-are running as so the `AI-assisted-by` trailer is set correctly.
-
-Do not push. Pushing is handled separately by the user.
+$ARGUMENTS
@@ -0,0 +1,10 @@
+---
+description: Investigate the chosen task, produce an implementation plan, and save it — loads and follows the make-a-plan skill
+agent: build
+---
+
+Load the **`make-a-plan`** skill and follow it as your only instruction.
+
+## User input, overrides and additional context
+
+$ARGUMENTS
@@ -1,40 +1,6 @@
 ---
-description: Resolve local git conflicts and stage the resolved files with git add — never continues the rebase
+description: Resolve local git conflicts and stage the resolved files; never continues the rebase — loads and follows the resolve-git-conflicts skill
 agent: build
 ---

-# Fix Git Conflicts
-
-Resolve conflicts in the local repository. The user handles finishing the
-rebase themselves — you must **never** run `git rebase --continue`,
-`git rebase --skip`, `git merge --continue`, or anything similar.
-
-## Phase 1 — Understand the problem (read-only)
-
-1. Run `git status` to detect the conflict state (rebase, merge, cherry-pick, etc.) and list conflicted files.
-2. For each conflicted (unmerged) file, understand the situation **without modifying anything**:
-   - Read the file and identify the conflict markers (`<<<<<<<`, `=======`, `>>>>>>>`).
-   - Inspect both sides — `git show <ours>:<file>` and `git show <theirs>:<file>` — plus `git log`/`git show` on the commits involved to understand intent.
-   - Identify what each side changed and why, and how they should be combined.
-
-## Phase 2 — Present the resolution plan
-
-3. **Present a clear plan to the user before touching any file.** For each conflicted file, state:
-   - What each side changed and why.
-   - Your proposed resolution and the reasoning behind it.
-   - How the two sides are combined (both additive → merge; both modify the same code → keep the semantically correct version, merging intent from both sides when clear from code and context).
-4. **Ask the user only when genuinely unclear.** Do not ask about anything you can determine yourself from the code, commit messages, or context. Only decisions that are not determinable and change the outcome (e.g. conflicting product decisions, which side to discard) warrant a question. **Collect all such questions together in an "Open Questions" section at the end of the plan**, so the user has full context to answer them properly.
-5. **Wait for the user to accept the plan** (and answer any open questions) before editing, staging, or otherwise modifying anything.
-
-## Phase 3 — Execute
-
-6. Resolve each conflicted file by editing the file to the agreed merged content and removing all conflict markers.
-
-## Phase 4 — Stage and verify
-
-7. **Stage every resolved file** with `git add <file>`. Do not stage unrelated untracked files unless clearly part of the resolution.
-8. Verify no conflict markers remain (search for `<<<<<<<` / `>>>>>>>` in resolved files) and that `git status` shows no unmerged paths.
-
-## Phase 5 — Report
-
-9. Briefly report the conflict state, how each conflicted file was resolved (and any answers received to open questions), and stop — do **not** run `git rebase --continue` or any other continuation command.
+Load the **`resolve-git-conflicts`** skill and follow it as your only instruction.
@@ -0,0 +1,10 @@
+---
+description: Code review — review a diff, PR, or code change — loads and follows the review-code skill
+agent: build
+---
+
+Load the **`review-code`** skill and follow it as your only instruction.
+
+## User input, overrides and additional context
+
+$ARGUMENTS
@@ -0,0 +1,10 @@
+---
+description: Plan review — evaluate an implementation plan before executing it — loads and follows the review-plan skill
+agent: build
+---
+
+Load the **`review-plan`** skill and follow it as your only instruction.
+
+## User input, overrides and additional context
+
+$ARGUMENTS
@@ -1,39 +0,0 @@
---
-name: create-pr
-description: Create or update a GitHub PR following Penpot conventions.
---
-
-# Skill: create-pr
-
-Create or update a GitHub PR. Read and follow:
- `mem:workflow/creating-prs` — title format, description structure, writing principles
- `mem:workflow/creating-commits` — commit type emojis
-
-## When to Use
-
- Creating a new PR from a feature branch
- Updating an existing PR's title or description to match conventions
-
-## Prerequisites
-
- `gh` CLI authenticated (`gh auth status`)
-
-## Commands
-
-**Create:**
-
-```bash
-gh pr create --repo penpot/penpot --title "<TITLE>" --body-file /tmp/pr-body.md
-```
-
-**Update:**
-
-```bash
-gh pr edit <NUMBER> --repo penpot/penpot --title "<TITLE>" --body-file /tmp/pr-body.md
-```
-
-**Verify:**
-
-```bash
-gh pr view <NUMBER> --repo penpot/penpot --json title,body
-```
@@ -1,377 +0,0 @@
---
-name: planner
-description: Read-only planning and architecture analysis for Penpot — produce a structured implementation plan with task breakdown, acceptance criteria, sizing, and checkpoints. Always output to the user and save to .opencode/plans/YYYY-MM-DD-<title>.md.
---
-
-# Planner
-
-Read-only senior software architect role for Penpot. Produces structured
-implementation plans with task breakdowns that engineers or other agents can
-execute. Never writes or modifies code.
-
-## When to Use
-
- The user asks for a plan, design, or analysis of a feature or bug.
- The user wants to understand which parts of the codebase a task will touch.
- The user needs a step-by-step implementation plan with file paths, function
-  names, and test strategy.
- The user asks "how would I implement X?" or "what's involved in fixing Y?".
- The user is about to start non-trivial work and wants a bite-sized task
-  breakdown.
- A task feels too large or vague to start.
- Work needs to be parallelized across multiple agents or sessions.
-
-Do **not** use this skill to actually implement anything — it is read-only.
-
-**When NOT to use:** Single-file changes with obvious scope, or when the spec
-already contains well-defined tasks.
-
-## Role
-
-You help users understand the Penpot codebase, design solutions, and produce
-implementation plans that other agents or developers can execute. The plan
-tells them what to build and how to verify it, task by task.
-
-The implementer reads the project's agent docs (`AGENTS.md`, project memories
-such as `mem:critical-info`, `mem:testing`, and each module's core memory)
-before working. Reference those memories instead of re-explaining tooling,
-conventions, or test design — explain in the plan only what they do not cover.
-
-Do **not** suggest commit messages or commit names anywhere in your plans or
-responses — committing is the implementer's responsibility.
-
-## CRITICAL: Required Reading Before Planning
-
-Before drafting any plan, work through the project's own guidance:
-
-1. Read `critical-info` (`.serena/memories/critical-info.md`) — the entry point
-   that describes the monorepo structure and module dependency graph.
-2. From `critical-info`, identify which modules your task affects.
-3. Read each affected module's core memory, e.g. `mem:frontend/core`,
-   `mem:backend/core`, `mem:common/core`, `mem:exporter/core`,
-   `mem:render-wasm/core`. Follow `mem:` references deeper as needed.
-4. For each affected module, note its lint, format, and test commands so the
-   plan can include concrete verification steps.
-
-Skipping this step is the #1 cause of incorrect or incomplete plans.
-
---
-
-## The Planning Process
-
-### Phase 1: Architecture Analysis
-
-1. Read the spec, requirements, or feature request.
-2. Analyze the codebase architecture and identify affected modules.
-3. Read project conventions (starting with `critical-info` and module core
-   memories) before drafting.
-4. Map dependencies between components (see the dependency graph in
-   `critical-info`).
-5. Identify risks, edge cases, performance implications, and breaking changes.
-
-### Phase 2: Task Breakdown
-
-#### Identify the Dependency Graph
-
-Map what depends on what, following the monorepo's module dependency graph:
-
-```
-common (shared types, schemas — no deps)
-    │
-    ├── backend (depends common)
-    │       ├── RPC handlers
-    │       └── persistence / migrations
-    │
-    ├── frontend (depends common, render-wasm)
-    │       ├── UI components
-    │       └── state / API integration
-    │
-    ├── exporter (depends common)
-    │
-    └── render-wasm (consumed by frontend)
-```
-
-Implementation order follows the dependency graph bottom-up: build shared
-foundations first, then layer consumers on top.
-
-#### Slice Vertically
-
-Instead of building all of common, then all of backend, then all of frontend —
-build one complete feature path at a time:
-
-**Bad (horizontal slicing):**
-```
-Task 1: Build all common types
-Task 2: Build all backend handlers
-Task 3: Build all frontend components
-```
-
-**Good (vertical slicing):**
-```
-Task 1: common data types + schema             ← foundation
-Task 2: backend RPC handler + persistence
-Task 3: frontend UI component + API integration
-```
-
-Each vertical slice delivers working, testable functionality.
-
-#### Write Tasks
-
-Each task follows this structure:
-
-```markdown
-## Task [N]: [Short descriptive title]
-
-**Description:** One or two paragraphs explaining what this task accomplishes.
-Should be clear and concise.
-
-**Rationale:** Why this task exists and why this approach over the obvious
-alternatives — design decisions, trade-offs, constraints discovered during
-analysis. One or two sentences; skip only if genuinely trivial.
-
-**Code sketch (optional):** Signature-, type-, or shape-level example when the
-intended interface is non-obvious. Keep it short — a skeleton that fixes the
-contract (function signature, model fields, error shape), never a full
-implementation. Omit when the task is mechanical.
-
-**Acceptance criteria:**
- [ ] [Specific, testable condition]
- [ ] [Specific, testable condition]
-
-**Verification:**
- [ ] Relevant tests pass (module-specific test command).
- [ ] Lint/formatter passes (module-specific check command), if applicable.
- [ ] The core flow works end-to-end, if applicable.
-
-**Dependencies:** [Task numbers this depends on, or "None"]
-
-**Files likely touched:**
- `path/to/file.clj`
- `path/to/file_test.clj`
-
-**Estimated scope:** [XS: 1 file | S: 1-2 files | M: 3-5 files | L: 5+ files]
-```
-
-Replace "module-specific test command" with the actual commands for the module
-(e.g. `clojure -M:dev:test` for backend/common,
-`npx shadow-cljs compile test && npx karma start` for frontend, or the
-commands noted in the module's core memory).
-
-When possible, design each task with TDD in mind: acceptance criteria double
-as a test list, and the natural first step of the task is writing those tests
-before the implementation. Some tasks resist this (config, migrations, pure
-wiring) — for those, keep the usual verification steps.
-
-#### Estimate Scope
-
-| Size | Files | Scope | Example |
-|------|-------|-------|---------|
-| **XS** | 1 | Single function, config change, or schema tweak | Add a validation rule |
-| **S** | 1-2 | One handler or component method | Add a new RPC endpoint |
-| **M** | 3-5 | One vertical feature slice | Bookmark CRUD with tests |
-| **L** | 5-8 | Multi-component feature | Search with filtering and pagination |
-| **XL** | 8+ | **Too large — break it down further** | — |
-
-If a task is XL, it should be broken into smaller tasks. Agents perform best
-on S and M tasks.
-
-**When to break a task down further:**
- It would take more than one focused session
- You cannot describe the acceptance criteria in 3 or fewer bullet points
- It touches two or more independent subsystems
- You find yourself writing "and" in the task title (a sign it is two tasks)
-
-#### Order and Checkpoints
-
-Arrange tasks so that:
-
-1. Dependencies are satisfied (build foundation first)
-2. Each task leaves the system in a working state
-3. Verification checkpoints occur after every 2-3 tasks
-4. High-risk tasks are early (fail fast)
-
-Add explicit checkpoints with the relevant module commands:
-
-```markdown
-### Checkpoint: After Tasks 1-3
- [ ] Relevant tests pass (module-specific command).
- [ ] The relevant build or compilation passes, if applicable.
- [ ] The core flow works end-to-end.
- [ ] Review with human before proceeding.
-```
-
-## Requirements
-
- Analyze the codebase architecture and identify affected modules.
- Read project conventions before drafting (start with `critical-info` and
-  affected module core memories).
- Break down complex features or bugs into atomic, actionable steps.
- Propose solutions with clear rationale, trade-offs, and sequencing.
- Identify risks, edge cases, performance implications, and breaking changes.
- Apply DRY and KISS principles to the proposed implementation.
- Define a testing strategy aligned with each affected module's tooling.
- Every task must have acceptance criteria and verification steps.
- Checkpoints must exist after every 2-3 tasks.
-
-## Constraints
-
- You are **analysis-only** — never create, edit, or delete source code.
- The only file write you may attempt is the plan itself, saved to
-  `.opencode/plans/`.
- You do **not** run builds, tests, linters, or any commands that modify state.
- You do **not** create git commits or interact with version control.
- You do **not** execute shell commands beyond read-only searches (`rg`, `ls`,
-  `find`, `cat`, `bat`).
- Your output is a structured plan or analysis, ready for handoff to an
-  engineer agent or developer.
-
-## Output Format
-
-The plan is always delivered in the response so the user sees it regardless
-of which agent is running the skill.
-
-Additionally, save the plan to:
-
-```
-.opencode/plans/YYYY-MM-DD-<plan-one-line-title>.md
-```
-
-Use today's date in the user's local timezone. The `<plan-one-line-title>`
-slug is lowercase, hyphen-separated, and a short summary of the task
-(e.g. `add-batch-get-profiles-for-file-comments`). Create the
-`.opencode/plans/` directory if it does not exist.
-
-IMPORTANT: The plan agent has write permission specifically for
-`.opencode/plans/` — always attempt the write. If the user explicitly provides
-a target file path, use that path instead of the default.
-
-### Plan Document Template
-
-```markdown
-# Plan: [Feature/Project Name]
-
-## Context
-[One paragraph: what is the problem or feature request? Why is it needed?]
-
-## Affected Modules
-[Which modules of the monorepo are involved? Reference module paths and any
-`mem:` memories that were consulted.]
-
-## Architecture Decisions
- [Key decision 1 and rationale]
- [Key decision 2 and rationale]
-
-## Risks & Considerations
-[Edge cases, performance implications, breaking changes, migration concerns,
-security implications.]
-
-## Approach
-[A short strategy summary: 3-5 sentences describing the overall approach and
-the shape of the dependency graph (what depends on what, what gets built
-first). High-level only — the task-by-task detail lives in the Task List.]
-
-## Task List
-
-Each task uses the full task structure defined in
-[Write Tasks](#write-tasks) — description, rationale, acceptance criteria,
-verification, dependencies, files, estimated scope, and optional code sketch.
-Never reduce a task to a one-line checkbox; the plan must be self-contained
-and executable without other context.
-
-Tasks are a flat, ordered list — a plan is not a roadmap. Do not group tasks
-into phases, milestones, or sprints; ordering and dependencies are already
-captured per task. Insert a checkpoint after every 2-3 tasks.
-
-## Task 1: [Short descriptive title]
-
-**Description:** [What this task accomplishes.]
-
-**Rationale:** [Why this approach over the alternatives.]
-
-**Acceptance criteria:**
- [ ] [Specific, testable condition]
-
-**Verification:**
- [ ] Relevant tests pass (module-specific command).
-
-**Dependencies:** None
-
-**Files likely touched:**
- `path/to/file`
-
-**Estimated scope:** [XS: 1 file | S: 1-2 files | M: 3-5 files | L: 5+ files]
-
-**Code sketch (optional):** [Short contract-level example, only if the shape
-is non-obvious.]
-
-## Task 2: [Short descriptive title]
-
-[Same structure as Task 1.]
-
-## Task 3: [Short descriptive title]
-
-[Same structure as Task 1.]
-
-### Checkpoint: After Tasks 1-3
- [ ] Relevant tests pass (module-specific command).
- [ ] The relevant build or compilation passes, if applicable.
- [ ] The core flow works end-to-end.
- [ ] Review with human before proceeding.
-
-## Task 4: [Short descriptive title]
-
-[Same structure as Task 1.]
-
-## Task 5: [Short descriptive title]
-
-[Same structure as Task 1.]
-
-## Verification & Testing
-[How to verify each task and the whole plan: the project's real test, lint,
-build, and run commands (extracted during Required Reading), coverage
-expectations, and manual checks. Consult each module's core memory for the
-exact commands.]
-
-## Parallelization Opportunities
- **Safe to parallelize:** Independent feature slices across separate
-  modules, tests for already-implemented features, documentation
- **Must be sequential:** Shared common schema changes, database migrations
- **Needs coordination:** Features that share a contract (define the contract
-  first, then parallelize)
-
-## Open Questions
- [Question needing human input]
-```
-
-When the plan is purely analytical (e.g. a code review or feasibility study
-with no implementation), skip the **Approach** and **Task List** sections and
-lead with **Findings** instead, keeping the rest of the structure.
-
-## Common Rationalizations
-
-| Rationalization | Reality |
-|---|---|
-| "I'll figure it out as I go" | That's how you end up with a tangled mess and rework. 10 minutes of planning saves hours. |
-| "The tasks are obvious" | Write them down anyway. Explicit tasks surface hidden dependencies and forgotten edge cases. |
-| "Planning is overhead" | Planning is the task. Implementation without a plan is just typing. |
-| "I can hold it all in my head" | Context windows are finite. Written plans survive session boundaries and compaction. |
-
-## Red Flags
-
- Delivering prose without a task breakdown
- Tasks that say "implement the feature" without acceptance criteria
- No verification steps in the plan
- All tasks are XL-sized
- No checkpoints between tasks
- Dependency order isn't considered
-
-## Verification Checklist
-
-Before delivering the plan, confirm:
-
- [ ] Every task has acceptance criteria
- [ ] Every task has a verification step
- [ ] Task dependencies are identified and ordered correctly
- [ ] No task is XL or larger — break it down instead
- [ ] Checkpoints exist after every 2-3 tasks
- [ ] The plan is ready for human review
@@ -11,10 +11,14 @@ You are working on the GitHub project `penpot/penpot`, a monorepo.
 # Development workflow

 - Commit/PR/issue creation is **on explicit request only**. Before any of these actions, read the relevant memory — don't infer format from prior examples:
-  - Before `git commit` → `mem:workflow/creating-commits` (subject format, body, `AI-assisted-by: model-name` trailer)
+  - Before `git commit` → `mem:workflow/creating-commits` (subject/body format, 76-char body wrapping enforced by `scripts/check-commit`, `AI-assisted-by: model-name` trailer)
  - Before `gh issue create` → `mem:workflow/creating-issues` (title derivation, body template, labels, Issue Type)
  - Before `gh pr create` / `gh pr edit` → `mem:workflow/creating-prs` (title format, body structure, "Note:" line)
+- Before a repo-wide pnpm version update → `mem:workflow/updating-pnpm` (workspace
+  layout, `corepack use` sweep order, the stamp-missing-field and
+  ignored-builds gotchas, verification steps)
 - **Never `git push`, force-push, or modify `git origin`** (or any other remote). The user pushes from their own shell; if a push is required, say so and wait. Never amend a commit that the user has already pushed unless explicitly asked.
+- **Never edit `CHANGES.md` by hand.** The changelog is generated from GitHub milestones during the release process; update it only via the `update-changelog` skill flow or on explicit user request.
 - You have access to the GitHub CLI `gh` or corresponding MCP tools.
 - Issues are also managed on Taiga. Read issues using the `read_taiga_issue` tool.
 - Before writing code, analyze the task in depth and describe your plan. If the task is complex, break it down into atomic steps.
@@ -70,6 +74,14 @@ module. You can read it from `mem:<MODULE>/core`
 - `scripts/error-reports.mjs` — Query error reports via RPC API with token
  authentication. Supports list/get operations with filtering and pagination.
  See `mem:scripts/error-reports`.
+- `scripts/clean-node-modules` — Remove stale `node_modules` from all pnpm
+  workspaces (root, modules, member packages). Keeps the shared pnpm store
+  at `<repo>/.pnpm-store` unless `--store`; ignores `external/` and
+  `.opencode/`. Usage and reinstall steps: `mem:workflow/updating-pnpm`.
+- `scripts/ci` — CI orchestration script: runs lint, tests, and format
+  checks per module (`frontend backend common render-wasm exporter mcp
+  plugins library`). Logs go to `.ci-logs/`; read the log file on failure.
+  See `mem:scripts/ci`.

 # Dependency graph

@@ -0,0 +1,61 @@
+# CI (scripts/ci)
+
+`scripts/ci` runs CI-style checks — lint, tests, format — for one or more
+monorepo modules and prints a per-task summary. It is the local equivalent
+of CI; use it to verify changes before declaring work done.
+
+## When to use
+
+- After implementing or fixing code in a module: run its checks before
+  finishing (AGENTS.md: run the applicable lint and format checks).
+- When `common/` changed: validate its consumers too (frontend, backend,
+  exporter; see the dependency graph in `mem:critical-info`).
+- To fix formatting across a module (`--fix`) or repair delimiters
+  (`--paren-repair`) before linting.
+
+## How to use (CLI)
+
+Run from the repo root:
+
+```bash
+./scripts/ci MODULE...                  # lint + test + fmt per module
+./scripts/ci --all --no-test            # lint + fmt on all modules
+./scripts/ci --lint frontend            # lint only
+./scripts/ci --fix --no-test frontend   # format files, skip tests
+./scripts/ci --paren-repair --all       # fix delimiters in all Clojure modules
+./scripts/ci --dry-run --all            # preview what would run
+```
+
+Modules: `frontend backend common render-wasm exporter mcp plugins library`.
+
+Flags:
+
+- Default tasks: `lint`, `test`, `fmt` (format check; `--fix` formats
+  instead).
+- `--lint` / `--test` / `--fmt` run one task only; `--no-lint` /
+  `--no-test` / `--no-fmt` drop one task from the default set.
+- `--paren-repair` runs only the delimiter repair — it wraps
+  `scripts/paren-repair` over each module's Clojure/CLJS sources; see
+  `mem:scripts/paren-repair`.
+- `--all` selects every module; `--exclude MOD` drops one (repeatable).
+- `--fail-fast` stops at the first failure; `--quiet` suppresses failure
+  output; `--dry-run` prints commands without running; `--clean` removes
+  the log directory.
+
+## Logs and exit codes
+
+- Full output of every task: `.ci-logs/<module>-<task>.log`.
+- On failure the script prints the last 30 lines; the final summary lists
+  every failed `module:task` with its log path.
+- Exit code 0 when all selected tasks passed, 1 otherwise.
+- Diagnose failures by reading the log file — never pipe test output
+  through filters (AGENTS.md hard rule).
+
+## Notes
+
+- `mcp` has no lint task (shows as skipped). `render-wasm` uses `./lint`,
+  `./test`, and `cargo fmt`.
+- Test tasks are long-running (backend: `clojure -M:dev:test`); use a
+  generous timeout when calling it from an agent shell.
+- Skill entry point: `.agents/skills/local-ci/SKILL.md`.
+- Testing principles and output discipline: `mem:testing`.
@@ -14,12 +14,32 @@ automatically pull the identity from the local git config `user.name` and `user.
 :emoji: Subject line (imperative, capitalized, no period, <=70 chars)

 Body explaining what changed and why.
-Wrap lines at 72 characters — git log and tooling
-render long lines poorly. Keep each line concise.
+Wrap lines at 76 characters — git log adds a
+four-space indent, so 76 + 4 fits an 80-column
+terminal. Keep each line concise.

 AI-assisted-by: model-name
 ```

+## HARD RULES (inexcusable)
+
+These rules are not advisory. Do not commit until every one holds. A commit
+that breaks them is wrong, even if the code is right.
+
+- **Body lines MUST wrap at 76 characters or fewer.** Measure every line; do
+  not eyeball it. This is the rule most often skipped. Rationale: `git log`
+  indents the body four spaces, so 76 + 4 fits an 80-column terminal.
+- **Subject MUST be ≤70 chars**, imperative, capitalized, no trailing period.
+- **MUST be a blank line** between subject and body.
+- **MUST run `scripts/check-commit` and get exit code 0 before finishing.**
+  It mechanically validates the rules above; a failing run is a blocker.
+  - It checks `HEAD` by default: `./scripts/check-commit`
+  - For another commit: `./scripts/check-commit -c <ref>`
+- **NEVER** hand-wave the body as "one long line". If a line exceeds 76,
+  break it at a space.
+- Exceptions inside the body (do not wrap these): `Signed-off-by:`,
+  `Co-authored-by:`, `AI-assisted-by:` trailers, and lines carrying a URL.
+
 **AI-assisted-by trailer rules:**
 - Use only the model name, e.g. `mimo-v2.5`, `deepseek-v4-flash`
 - Do NOT add prefixes like `opencode-go/` — use the bare model name
@@ -351,5 +351,5 @@ gh issue view <NUMBER> --repo penpot/penpot --json title
 ## See Also

 - End-to-end orchestration entry point: the `create-issue` skill at
-  `.opencode/skills/create-issue/SKILL.md`. The skill is a thin entry
+  `.agents/skills/create-issue/SKILL.md`. The skill is a thin entry
  point; this memory is the canonical home for all issue-creation rules.
@@ -0,0 +1,91 @@
+# Updating pnpm Across All Workspaces
+
+Canonical procedure. Run it from the repo root with the log redirected to a
+file (never pipe tool output through filters).
+
+## Layout facts
+
+- The repo has 11 pnpm workspaces, each with its own `pnpm-workspace.yaml`
+  and `pnpm-lock.yaml`: the repo root plus `backend`, `common`, `docs`,
+  `exporter`, `frontend`, `library`, `mcp`, `media-processor`, `plugins`,
+  and `render-wasm`.
+- Every package inside a module workspace (for example all `plugins/apps/*`
+  and `plugins/libs/*` packages) is a plain member of that module's
+  workspace. Members must not carry their own `pnpm-workspace.yaml` or
+  `pnpm-lock.yaml`; their dependencies resolve through the parent
+  workspace's lockfile.
+- One shared pnpm store for the whole repo: `<repo>/.pnpm-store`. Every
+  workspace yaml sets it explicitly: `storeDir: .pnpm-store` at the root,
+  `storeDir: ../.pnpm-store` in each module. pnpm resolves the value
+  against the workspace root, so all workspaces land on the same store.
+  Do not remove these lines: nested workspaces do not inherit settings,
+  and without them each workspace may resolve a different store.
+- The store survives `node_modules` cleans. It is content-addressed and
+  integrity-verified, so it cannot go stale; staleness lives in
+  node_modules. Only `scripts/clean-node-modules --store` removes it.
+- Every `package.json` (about 35 of them) must carry a `packageManager` field
+  with the identical `pnpm@<version>+sha512.<hash>` value. Do not let them drift.
+- CI pins no pnpm version; workflows rely on corepack reading
+  `packageManager`. Fixing the fields fixes CI.
+
+## Procedure
+
+1. Resolve the target tag first and note the version. Example:
+   `npm view pnpm dist-tags --json` for `next-12` (latest 12.x). The tag
+   moves over time; always re-check.
+2. List every directory with a `package.json`, excluding `node_modules`
+   (`fd -H -t f package.json -E node_modules`). This list is the work set;
+   do not maintain a hand-written list.
+3. Run `corepack use pnpm@<tag>` in workspace roots first, then members.
+   `corepack use` stamps `packageManager` in the nearest package.json and
+   runs an install. Member runs repeat the workspace install; after the root
+   run they are quick no-ops.
+4. If a run fails, fix the cause (see gotchas) and re-run that directory.
+
+## Gotchas
+
+- `corepack use` only updates an existing `packageManager` field. If a
+  package.json lacks the field, corepack walks up to the nearest ancestor
+  that has one and stamps that file instead; the member stays unstamped.
+  After the sweep, assert every package.json carries the field. For a
+  missing one, insert the identical `pnpm@<version>+sha512.<hash>` string,
+  then re-run `corepack use pnpm@<tag>` in that directory.
+- A workspace may fail with `ERR_PNPM_IGNORED_BUILDS`, and pnpm then writes
+  a placeholder scaffold into its `pnpm-workspace.yaml`:
+  `allowBuilds: esbuild: set this to true or false`. Current pnpm writes
+  only the `allowBuilds` placeholder; any legacy key still present
+  (`ignoredBuiltDependencies`, `onlyBuiltDependencies`,
+  `neverBuiltDependencies`) is ignored since pnpm 11. Repo convention is
+  `allowBuilds: esbuild: true`. Replace the placeholder and drop the
+  legacy entry, then re-run.
+- `plugins/apps/composable-test-suite` once had its own
+  `pnpm-workspace.yaml` and acted as a nested workspace root. That state is
+  gone on purpose: pnpm picks the nearest `pnpm-workspace.yaml` walking up,
+  so a nested one silently forks install and lockfile behavior. Do not
+  reintroduce it.
+- Expect metadata-only lockfile diffs when only the pnpm version moves:
+  the pnpm self-reference entries, plus a new `packageManagerDependencies`
+  section in lockfiles last written by older pnpm. Large diffs mean
+  re-resolution; inspect them before accepting.
+
+## Verification
+
+- Every `packageManager` field is byte-identical (same version and hash).
+- `pnpm --version` in each workspace prints the target version.
+- `pnpm install --frozen-lockfile` succeeds in each of the 11 workspaces.
+- `git diff` on lockfiles matches the expectations above.
+
+## Cleaning stale node_modules
+
+- `scripts/clean-node-modules` removes every workspace `node_modules`: the
+  repo root, all module workspaces, and all member packages. Use it when
+  installs misbehave after dependency changes: clean, reinstall, done.
+- Flags: `-n/--dry-run` lists without deleting; `--store` also removes the
+  shared pnpm store at `<repo>/.pnpm-store` (the next install re-downloads
+  what it held). `external/` (vendored dependency trees with their own
+  lifecycles) and `.opencode/` are always ignored.
+- The script never touches the pnpm store by default, so the reinstall
+  after cleaning reuses cached packages (zero downloads).
+- After cleaning, run `pnpm install` in each workspace root to restore the
+  development environment; `frontend` postinstall also reinstalls and
+  builds `plugins-runtime`.
@@ -8,9 +8,18 @@
  wait for the user to push. Do not change the remote URL, do not switch SSH↔HTTPS.
 - **Never amend a commit that has been pushed** unless the user explicitly asks.
  If the user pushes, treat that commit as final from the agent's side.
+- **Never edit `CHANGES.md` by hand** in commits or PRs. The changelog is
+  generated from GitHub milestones during the release process; update it only
+  via the `update-changelog` skill flow or on explicit user request.
 - **Never pipe test output directly to filters** (`| head`, `| tail`, `| grep`, etc.).
  Always redirect to a file first: `command > /tmp/output.txt 2>&1`, then read/grep the file.
  This prevents hiding test failures. See `mem:testing` for details.
+- **`.claude/skills` is a symlink to `.agents/skills`.**
+  Edit skills only in their canonical location (`.agents/skills`); never edit
+  through `.claude/skills`.
+- **Commit message body lines MUST wrap at ≤76 chars** (subject ≤70 chars) and
+  the commit MUST pass `./scripts/check-commit` with exit code 0 before you
+  consider it done. This is mechanically checked — do not eyeball it.
 - **Read the workflow memory BEFORE the corresponding action**:
  - Before `git commit` → `mem:workflow/creating-commits` (commit format, AI-assisted-by trailer)
  - Before `gh issue create` → `mem:workflow/creating-issues` (title derivation, body template, Issue Type)
@@ -142,6 +151,6 @@ precision while maintaining a strong focus on maintainability and performance.
 - `scripts/nrepl-eval.mjs` — Evaluate Clojure code via nREPL (backend + frontend).
 - `scripts/check-commit` — Validate commit messages against Penpot's commit guidelines.
 - `scripts/check-fmt-clj` — Check Clojure formatting without modifying files.
- `scripts/ci` — CI orchestration script for running lint, tests, and format checks across modules. See `scripts/ci --help`.
+- `scripts/ci` — CI orchestration script for running lint, tests, and format checks across modules. See `mem:scripts/ci`.
 - `scripts/gh.py` — Multi-purpose GitHub CLI helper. Subcommands: `issues` (list issues in a milestone), `prs` (fetch PR details), `advisories` (list/inspect security advisories). See `python3 scripts/gh.py --help`.

@@ -2,6 +2,12 @@

 ## 2.19.0 (Unreleased)

+### :rocket: Epics and highlights
+
+- Add configurable keyboard shortcuts [#9924](https://github.com/penpot/penpot/issues/9924) (PR: [#10237](https://github.com/penpot/penpot/pull/10237))
+- Improve path operations and edition in the path editor [#10889](https://github.com/penpot/penpot/issues/10889) (PR: [#10807](https://github.com/penpot/penpot/pull/10807))
+- Add auto-linking of libraries during import based on slugified name [#9263](https://github.com/penpot/penpot/issues/9263) (PR: [#9958](https://github.com/penpot/penpot/pull/9958))
+
 ### :bug: Bugs fixed

 - Fix copying text from Penpot to the clipboard not working on MS Windows [#11303](https://github.com/penpot/penpot/issues/11303) (PR: [#11305](https://github.com/penpot/penpot/pull/11305))
@@ -11,6 +17,20 @@
 - Fix overlay shifting left when shown with top-center alignment in viewer prototype (by @filipsajdak) [#9048](https://github.com/penpot/penpot/issues/9048) (PR: [#10454](https://github.com/penpot/penpot/pull/10454))
 - Fix internal error when clicking the Copy button on the Access Token page (by @0xTHAC0) [#8496](https://github.com/penpot/penpot/issues/8496) (PR: [#11156](https://github.com/penpot/penpot/pull/11156))
 - Fix `disable-registration` flag not preventing non-users from creating accounts in the share prototypes page (by @0xTHAC0) [#5164](https://github.com/penpot/penpot/issues/5164) (PR: [#11199](https://github.com/penpot/penpot/pull/11199))
+- Fix "Cannot assign to read only property 'toString'" error during text resize (by @makesomethingshit) [#10168](https://github.com/penpot/penpot/issues/10168) (PR: [#11521](https://github.com/penpot/penpot/pull/11521))
+- Fix plugin postMessage channel broadcasting messages to all plugins without origin validation [#10968](https://github.com/penpot/penpot/issues/10968) (PR: [#10970](https://github.com/penpot/penpot/pull/10970))
+- Fix MCP plugin page navigation while connected crashing the workspace (by @makesomethingshit) [#11001](https://github.com/penpot/penpot/issues/11001) (PR: [#11521](https://github.com/penpot/penpot/pull/11521))
+- Fix shortcut search never matching on key combination, only on action label [#11003](https://github.com/penpot/penpot/issues/11003) (PR: [#11081](https://github.com/penpot/penpot/pull/11081))
+- Fix Shift + special character key shortcut capturing the shifted character instead of the physical key [#11004](https://github.com/penpot/penpot/issues/11004) (PR: [#11081](https://github.com/penpot/penpot/pull/11081))
+- Fix reassigning the "Paste" shortcut not updating the UI or taking effect in the workspace [#11005](https://github.com/penpot/penpot/issues/11005) (PR: [#11081](https://github.com/penpot/penpot/pull/11081))
+- Fix font-size dropdown clipping multi-digit values in Firefox (by @0xTHAC0) [#11008](https://github.com/penpot/penpot/issues/11008) (PR: [#11162](https://github.com/penpot/penpot/pull/11162), [#11500](https://github.com/penpot/penpot/pull/11500))
+- Fix exporting shortcuts producing an invalid "toggle-fullscreen" entry that breaks re-import [#11032](https://github.com/penpot/penpot/issues/11032) (PR: [#11081](https://github.com/penpot/penpot/pull/11081))
+- Fix plugin API missing permission checks in tokens, shapes, variants, flows, layouts, and user identity [#11137](https://github.com/penpot/penpot/issues/11137) (PR: [#11139](https://github.com/penpot/penpot/pull/11139))
+- Fix library summary Redis cache keys omitting the tenant [#11407](https://github.com/penpot/penpot/issues/11407) (PR: [#11408](https://github.com/penpot/penpot/pull/11408))
+- Fix active theme name in the inspect tab displaying an id instead of the name [#11437](https://github.com/penpot/penpot/issues/11437) (PR: [#11439](https://github.com/penpot/penpot/pull/11439))
+- Fix triple-click not selecting the full line in text editor v3 [#11483](https://github.com/penpot/penpot/issues/11483) (PR: [#11493](https://github.com/penpot/penpot/pull/11493))
+- Fix pasted text losing formatting on last lines after resizing and adding new lines from the top [#11501](https://github.com/penpot/penpot/issues/11501) (PR: [#11503](https://github.com/penpot/penpot/pull/11503))
+- Fix variant property dropdown appearing empty and throwing an internal error when the component has no sibling variants [#11524](https://github.com/penpot/penpot/issues/11524) (PR: [#11499](https://github.com/penpot/penpot/pull/11499))

 ### :sparkles: New features & Enhancements

@@ -19,9 +39,22 @@
 - Improve path operations and edition in the path editor [#10889](https://github.com/penpot/penpot/issues/10889) (PR: [#10807](https://github.com/penpot/penpot/pull/10807))
 - Add configurable keyboard shortcuts [#9924](https://github.com/penpot/penpot/issues/9924) (PR: [#10237](https://github.com/penpot/penpot/pull/10237))
 - Add auto-linking of libraries during import based on slugified name [#9263](https://github.com/penpot/penpot/issues/9263) (PR: [#9958](https://github.com/penpot/penpot/pull/9958))
+- Add support for internal libraries and file sync for Design Tokens [#9334](https://github.com/penpot/penpot/issues/9334)
+- Warn self-hosted users when their Penpot version is outdated and surface what they're missing [#10497](https://github.com/penpot/penpot/issues/10497) (PR: [#11411](https://github.com/penpot/penpot/pull/11411))
+- Add dedicated RPC methods for plugin registry operations with permission validation [#10952](https://github.com/penpot/penpot/issues/10952) (PR: [#10957](https://github.com/penpot/penpot/pull/10957))
+- Document MCP and internal resolver environment variables (by @ShreyashAgare26) [#11318](https://github.com/penpot/penpot/issues/11318) (PR: [#11572](https://github.com/penpot/penpot/pull/11572))
+- Add tokens source indicator to assets tab [#11365](https://github.com/penpot/penpot/issues/11365) (PR: [#11439](https://github.com/penpot/penpot/pull/11439))
+- Export multiple fills to SVG [#11466](https://github.com/penpot/penpot/issues/11466) (PR: [#11467](https://github.com/penpot/penpot/pull/11467))
+- Add Penpot-specific board size presets (file thumbnail, template cover, plugin icon/cover) [#11561](https://github.com/penpot/penpot/issues/11561) (PR: [#11565](https://github.com/penpot/penpot/pull/11565))

 ## 2.18.0 (Unreleased)

+### :rocket: Epics and highlights
+
+- Group toolbar drawing tools into shape and free-draw flyouts [#9316](https://github.com/penpot/penpot/issues/9316) (PR: [#9480](https://github.com/penpot/penpot/pull/9480), [#10354](https://github.com/penpot/penpot/pull/10354))
+- Add dedicated Line and Arrow drawing tools (by @davidv399) [#9145](https://github.com/penpot/penpot/issues/9145) (PR: [#9146](https://github.com/penpot/penpot/pull/9146))
+- Show and manage comments while designing in the workspace [#10239](https://github.com/penpot/penpot/issues/10239) (PR: [#10275](https://github.com/penpot/penpot/pull/10275))
+
 ### :bug: Bugs fixed

 - Fix MCP integration hanging when the Penpot tab is backgrounded or frozen by the browser [#10323](https://github.com/penpot/penpot/issues/10323) (PR: [#10392](https://github.com/penpot/penpot/pull/10392))
@@ -123,6 +156,21 @@
 - Fix incorrect permission handling when managing share links on a file [#11289](https://github.com/penpot/penpot/issues/11289) (PR: [#11290](https://github.com/penpot/penpot/pull/11290))
 - Fix backend session remaining valid after logout when the auth-token cookie is replayed [#11316](https://github.com/penpot/penpot/issues/11316) (PR: [#11317](https://github.com/penpot/penpot/pull/11317))
 - Fix get-team-invitation-token requiring only read permissions [#11358](https://github.com/penpot/penpot/issues/11358) (PR: [#11359](https://github.com/penpot/penpot/pull/11359))
+- Fix missing text in legacy SVG board thumbnails [#10182](https://github.com/penpot/penpot/issues/10182) (PR: [#11552](https://github.com/penpot/penpot/pull/11552))
+- Fix workspace crash when applying transform modifiers in the WASM renderer [#10894](https://github.com/penpot/penpot/issues/10894) (PR: [#10896](https://github.com/penpot/penpot/pull/10896))
+- Limit ZIP entry count and object size on V3 binfile import [#11021](https://github.com/penpot/penpot/issues/11021) (PR: [#11022](https://github.com/penpot/penpot/pull/11022))
+- Block plugin UI iframe URLs targeting the Penpot domain [#11271](https://github.com/penpot/penpot/issues/11271) (PR: [#11273](https://github.com/penpot/penpot/pull/11273))
+- Restrict the MCP REPL code execution endpoint to development environments [#11283](https://github.com/penpot/penpot/issues/11283) (PR: [#11282](https://github.com/penpot/penpot/pull/11282))
+- Filter share-link tokens from the get-view-only-bundle response [#11285](https://github.com/penpot/penpot/issues/11285) (PR: [#11286](https://github.com/penpot/penpot/pull/11286))
+- Disable MCP developer tools in multi-user mode [#11291](https://github.com/penpot/penpot/issues/11291) (PR: [#11310](https://github.com/penpot/penpot/pull/11310))
+- Fix Hide comments setting being ignored after opening the Comments section [#11308](https://github.com/penpot/penpot/issues/11308) (PR: [#11492](https://github.com/penpot/penpot/pull/11492))
+- Block NAT64/6to4/Teredo IPv6 transition addresses in the SSRF guard [#11319](https://github.com/penpot/penpot/issues/11319) (PR: [#11320](https://github.com/penpot/penpot/pull/11320))
+- Prevent team admins from removing the team owner [#11367](https://github.com/penpot/penpot/issues/11367) (PR: [#11368](https://github.com/penpot/penpot/pull/11368))
+- Enforce share-link comment permissions and page scope [#11370](https://github.com/penpot/penpot/issues/11370) (PR: [#11371](https://github.com/penpot/penpot/pull/11371))
+- Clean up orphaned teams, projects and files on profile deletion [#11394](https://github.com/penpot/penpot/issues/11394) (PR: [#11395](https://github.com/penpot/penpot/pull/11395))
+- Fix crash when pressing Ctrl+D with no shape selected [#11448](https://github.com/penpot/penpot/issues/11448) (PR: [#11491](https://github.com/penpot/penpot/pull/11491))
+- Fix text layout not updating when auto-width is set by double-clicking the bounding box [#11480](https://github.com/penpot/penpot/issues/11480) (PR: [#11541](https://github.com/penpot/penpot/pull/11541))
+- Fix boolean shapes rendering deformed in the WASM renderer and exports [#11482](https://github.com/penpot/penpot/issues/11482) (PR: [#11551](https://github.com/penpot/penpot/pull/11551))

 ### :sparkles: New features & Enhancements

@@ -185,6 +233,10 @@
 ### :rocket: Epics and highlights

 - Render prototype viewer with WASM (Skia) engine instead of SVG [#10037](https://github.com/penpot/penpot/issues/10037) (PR: [#10038](https://github.com/penpot/penpot/pull/10038))
+- Add layer blur effect for visual depth and styling [#9844](https://github.com/penpot/penpot/issues/9844) (PR: [#10034](https://github.com/penpot/penpot/pull/10034))
+- Render guides in WebGL for consistent viewer performance [#10068](https://github.com/penpot/penpot/issues/10068) (PR: [#10014](https://github.com/penpot/penpot/pull/10014))
+- Add concurrency limiter and status indicators for MCP server communications [#9493](https://github.com/penpot/penpot/issues/9493) (PR: [#9748](https://github.com/penpot/penpot/pull/9748))
+- Add typography token row to multiselected texts for better token visibility [#9336](https://github.com/penpot/penpot/issues/9336) (PR: [#9128](https://github.com/penpot/penpot/pull/9128))

 ### :sparkles: New features & Enhancements

@@ -360,7 +412,7 @@

 ### :rocket: Epics and highlights

- WebGL rendering (beta) user preference [#9683](https://github.com/penpot/penpot/issues/9683) (PR:[9113](https://github.com/penpot/penpot/pull/9113))
+- WebGL rendering (beta) user preference [#9683](https://github.com/penpot/penpot/issues/9683) (PR: [#9113](https://github.com/penpot/penpot/pull/9113))
 - Design Tokens at the design tab: numeric fields with token selection in place [#9358](https://github.com/penpot/penpot/issues/9358)

 ### :sparkles: New features & Enhancements
@@ -539,6 +591,10 @@

 ## 2.15.0

+### :rocket: Epics and highlights
+
+- Add MCP server integration for AI-assisted design workflows [#9174](https://github.com/penpot/penpot/issues/9174) (PR: [#9032](https://github.com/penpot/penpot/pull/9032), [#9321](https://github.com/penpot/penpot/pull/9321))
+
 ### :sparkles: New features & Enhancements

 - Add MCP server integration [GH #9174](https://github.com/penpot/penpot/issues/9174)
@@ -0,0 +1,3 @@
+Read and follow the instructions in `AGENTS.md`.
+
+Treat `AGENTS.md` as the canonical project instruction file.
@@ -188,8 +188,11 @@ Commit messages must follow this format:
 - Add clear and concise description on the body
 - Do not end the subject with a period
 - Keep the subject to **70 characters** or fewer
+- **Wrap body lines at 76 characters or fewer** (trailers and URLs excepted)
 - Separate the subject from the body with a **blank line**

+You can check a commit against these rules with `./scripts/check-commit`.
+
 ### Examples

 ```
@@ -1,26 +0,0 @@
-# HIGHLIGHTS
-
-## 2.17.0
-
- Background blur is here
- WebGL rendering gets stronger
- MCP connection status and more
- Design tokens: more visible, more user-friendly
-
-
-## 2.16.0
-
- Design tokens in the design panel
- Major community contributions
- WebGL rendering (beta)
-
-
-## 2.15.0
-
- AI connected to real design context
- Multi-directional workflow
- Your stack, your model, your decision
-
-
-
-
@@ -56,6 +56,7 @@ If your organization is scaling and needs extra support, we’re here to help. [

 - [Why Penpot](#why-penpot)
 - [Getting Started](#getting-started)
+- [Penpot Enterprise](#penpot-enterprise)
 - [Community](#community)
 - [Contributing](#contributing)
 - [Resources](#resources)
@@ -93,6 +94,12 @@ Penpot is the only design & prototype platform that is deployment agnostic. You

 Learn how to install it with Docker, Kubernetes, Elestio or other options on [our website](https://penpot.app/self-host).

+<img width="100%" height="1010" alt="2" src="https://github.com/user-attachments/assets/243e796e-a140-481a-b68f-b24be6a70e37" />
+
+## Penpot Enterprise ##
+
+Penpot Enterprise is our paid plan for organizations that need to scale their design work across multiple teams with advanced governance, security, and administration. Manage teams and access from a centralized **Admin Console**, configure advanced permissions, and connect your **identity provider through SSO**. Available for cloud and self-hosted environments, it combines enterprise controls with Penpot’s open-source foundation and open standards.
+
 ## Community ##

 We love the Open Source software community. Contributing is our passion and if it’s yours too, participate and [improve](https://community.penpot.app/c/help-us-improve-penpot/7) Penpot. All your designs, code and ideas are welcome!
@@ -4,7 +4,7 @@
  "license": "MPL-2.0",
  "author": "Kaleidos INC Sucursal en España SL",
  "private": true,
-  "packageManager": "pnpm@11.20.0+sha512.9a6f330a95b66446ea088faf1521405a8a01f07fde7124cc9958dfed52d4bb436737e65b08f85f37b46fcba375092558ac51262b816844b22f63406ed166bfee",
+  "packageManager": "pnpm@12.3.4+sha512.961aa41fb077da3a04a441d9f8e15ebc0c96da8ef710b2eb67bf9ee7cb0610eabd48f1fd85f51cffe73846785fa0f87c56a3a872a1d893f8446741b5cce45457",
  "repository": {
    "type": "git",
    "url": "https://github.com/penpot/penpot"
@@ -1,3 +1,104 @@
+---
+lockfileVersion: '9.0'
+
+importers:
+
+  .:
+    configDependencies: {}
+    packageManagerDependencies:
+      pnpm:
+        specifier: 12.3.4
+        version: 12.3.4
+
+packages:
+
+  '@pnpm/exe.darwin-arm64@12.3.4':
+    resolution: {integrity: sha512-PAyUol8T1+/+ViOiXAt51ECA+QnfXCqz6foL4bW+LsoX0NcVd5XVEM2mRQu+LV4oc7uRz9zf9U0P+XFfuQeDAw==}
+    cpu: [arm64]
+    os: [darwin]
+
+  '@pnpm/exe.darwin-x64@12.3.4':
+    resolution: {integrity: sha512-fxP9JCk0Cdye+ePuj+GJJLMUMTqHGWRdb1dtv4How876uQ2ehxvenpgiYAir/ceO9PsYUZkFTtyZdx+rRu5QOA==}
+    cpu: [x64]
+    os: [darwin]
+
+  '@pnpm/exe.linux-arm64-musl@12.3.4':
+    resolution: {integrity: sha512-FBOt0/7ye6O6q4AllVV5QMviB6qE6fqkeczV/+MDWQsmo+QJrlfsh6X7CpH/tClVpBZEyIbjpUoT8bNhCYBxEg==}
+    cpu: [arm64]
+    os: [linux]
+    libc: [musl]
+
+  '@pnpm/exe.linux-arm64@12.3.4':
+    resolution: {integrity: sha512-t71AVA7LRqiKTyZ5xMYaZc2n5DfdpMbfokZuiIOXHBOM03ECnF0t4iYwaBDqJgVjlKYUOwaF/bRQajGNA4cJ4w==}
+    cpu: [arm64]
+    os: [linux]
+    libc: [glibc]
+
+  '@pnpm/exe.linux-x64-musl@12.3.4':
+    resolution: {integrity: sha512-RPmk7Jb/aYaFvL2iyDN/AtMY+hUEsue732WmXpcuQ9tBpMnGyA5py7Z3+e+qmQaJ0zY/4ni9jJiyPBQHujmv6w==}
+    cpu: [x64]
+    os: [linux]
+    libc: [musl]
+
+  '@pnpm/exe.linux-x64@12.3.4':
+    resolution: {integrity: sha512-2ZqOlSPkfwX1h5cR+FPiWf8+F+2hZT/3TvhUK5sigHqwaQCIiq8R7CGxhndKs63JtcLi2a1Qpo+wX/EoyfjyJQ==}
+    cpu: [x64]
+    os: [linux]
+    libc: [glibc]
+
+  '@pnpm/exe.win32-arm64@12.3.4':
+    resolution: {integrity: sha512-ANyrHqyqco6SXBysUTRF74itDyyraea7IbFsKFdNXTjcFnfycTDx37EwuhdpPYFNSIh2JhUG4fByclsRfiHX7w==}
+    cpu: [arm64]
+    os: [win32]
+
+  '@pnpm/exe.win32-x64@12.3.4':
+    resolution: {integrity: sha512-WH/KqBPY/hq2Tb7SgQltEZytimcjgKRaCRL/aM9CI0c67iKc5TVmHUhIiL3Ux9FB4bWn36i6XewUcScQI+zG8w==}
+    cpu: [x64]
+    os: [win32]
+
+  pnpm@12.3.4:
+    resolution: {integrity: sha512-lhqkH7B32joEpEHZ+OFevAyW2o73ELLrZ7+e58sGEOq9SPH9hfUc/+c4RnhfoPh8VqOocqHYk/hEZ0G1zORUVw==}
+    engines: {node: '>=18.*'}
+    hasBin: true
+
+snapshots:
+
+  '@pnpm/exe.darwin-arm64@12.3.4':
+    optional: true
+
+  '@pnpm/exe.darwin-x64@12.3.4':
+    optional: true
+
+  '@pnpm/exe.linux-arm64-musl@12.3.4':
+    optional: true
+
+  '@pnpm/exe.linux-arm64@12.3.4':
+    optional: true
+
+  '@pnpm/exe.linux-x64-musl@12.3.4':
+    optional: true
+
+  '@pnpm/exe.linux-x64@12.3.4':
+    optional: true
+
+  '@pnpm/exe.win32-arm64@12.3.4':
+    optional: true
+
+  '@pnpm/exe.win32-x64@12.3.4':
+    optional: true
+
+  pnpm@12.3.4:
+    optionalDependencies:
+      '@pnpm/exe.darwin-arm64': 12.3.4
+      '@pnpm/exe.darwin-x64': 12.3.4
+      '@pnpm/exe.linux-arm64': 12.3.4
+      '@pnpm/exe.linux-arm64-musl': 12.3.4
+      '@pnpm/exe.linux-x64': 12.3.4
+      '@pnpm/exe.linux-x64-musl': 12.3.4
+      '@pnpm/exe.win32-arm64': 12.3.4
+      '@pnpm/exe.win32-x64': 12.3.4
+
+---
 lockfileVersion: '9.0'

 settings:
@@ -1,2 +1,4 @@
+storeDir: ../.pnpm-store
+
 minimumReleaseAgeExclude:
  - brace-expansion@5.0.8 || 5.0.9
@@ -31,7 +31,7 @@ export PENPOT_MEDIA_PROCESSING_SERVICE_URI=http://localhost:6065
 export PENPOT_FLAGS="\
       $PENPOT_FLAGS \
       enable-login-with-password \
-       disable-login-with-ldap \
+       enable-login-with-ldap \
       disable-login-with-oidc \
       disable-login-with-google \
       disable-login-with-github \
@@ -10,7 +10,7 @@
   [app.common.logging :as l]
   [app.common.schema :as sm]
   [clj-ldap.client :as ldap]
-   [clojure.string]
+   [cuerdas.core :as str]
   [integrant.core :as ig]))

 (defn- prepare-params
@@ -36,11 +36,22 @@
                :cause cause))))

 (defn- replace-several [s & {:as replacements}]
-  (reduce-kv clojure.string/replace s replacements))
+  (reduce-kv str/replace s replacements))
+
+(defn- escape-ldap-filter-value
+  "Escapes special characters in a string for use in LDAP filter values,
+  per RFC 4515 section 3."
+  [s]
+  (-> s
+      (str/replace "\\" "\\5c")
+      (str/replace "*"  "\\2a")
+      (str/replace "("  "\\28")
+      (str/replace ")"  "\\29")
+      (str/replace "\u0000" "\\00")))

 (defn- search-user
  [{:keys [::conn base-dn] :as cfg} email]
-  (let [query  (replace-several (:query cfg) ":username" email)
+  (let [query  (replace-several (:query cfg) ":username" (escape-ldap-filter-value email))
        attrs  [(:attrs-username cfg)
                (:attrs-email cfg)
                (:attrs-fullname cfg)]
@@ -49,12 +60,19 @@
                 :attributes attrs}]
    (first (ldap/search conn base-dn params))))

+(defn- get-attr
+  "Retrieves an attribute from an LDAP entry. Handles multi-valued
+  attributes by returning the first value."
+  [entry attr-key]
+  (let [v (get entry attr-key)]
+    (if (coll? v) (first v) v)))
+
 (defn- retrieve-user
  [{:keys [::conn] :as cfg} {:keys [email password]}]
  (when-let [{:keys [dn] :as user} (search-user cfg email)]
    (when (ldap/bind? conn dn password)
-      {:fullname (get user (-> cfg :attrs-fullname keyword))
-       :email    email
+      {:fullname (get-attr user (-> cfg :attrs-fullname keyword))
+       :email    (get-attr user (-> cfg :attrs-email keyword))
       :backend  "ldap"})))

 (def ^:private schema:info-data
@@ -79,7 +97,7 @@
          (l/warn :hint "invalid response from ldap, looks like ldap is not configured correctly" :data user)
          (ex/raise :type :restriction
                    :code :wrong-ldap-response
-                    :explain explain)))
+                    ::sm/explain explain)))
      user)))

 (defn- try-connectivity
@@ -1037,7 +1037,7 @@
            provider        (prepare-organization-sso-provider cfg sso)
            _info           (get-info cfg provider state code)
            session         (session/get-session request)
-            exp             (ct/in-future {:minutes 15})]
+            exp             (ct/in-future {:hours 4})]
        (when (and session organization-id)
          (let [props (-> (or (:props session) {})
                          (update :sso assoc organization-id exp))]
@@ -58,6 +58,7 @@
   :objects-storage-fs-directory "assets"

   :auth-token-cookie-name "auth-token"
+   :auth-token-cookie-max-age-absolute (ct/duration {:days 30})

   :assets-path "/internal/assets/"
   :smtp-default-reply-to "Penpot <no-reply@example.com>"
@@ -91,6 +92,7 @@

   :quotes-upload-sessions-per-profile 5
   :quotes-upload-chunks-per-session 20
+   :upload-max-chunk-size (* 1024 1024 30) ; 30MiB

   ;; SSRF protection
   :ssrf-allowed-hosts #{}
@@ -202,10 +204,12 @@
    [:quotes-team-access-requests-per-requester {:optional true} ::sm/int]
    [:quotes-upload-sessions-per-profile {:optional true} ::sm/int]
    [:quotes-upload-chunks-per-session {:optional true} ::sm/int]
+    [:upload-max-chunk-size {:optional true} ::sm/int]
    [:quotes-media-storage-bytes-per-team {:optional true} ::sm/int]

    [:auth-token-cookie-name {:optional true} :string]
    [:auth-token-cookie-max-age {:optional true} ::ct/duration]
+    [:auth-token-cookie-max-age-absolute {:optional true} ::ct/duration]

    [:registration-domain-whitelist {:optional true} [::sm/set :string]]
    [:email-verify-threshold {:optional true} ::ct/duration]
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.arrow
  "Bulk Ladybug ingest through in-memory Arrow.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.debug
  "In-memory Ladybug sessions for the debug graph console."
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.ingest
  "Penpot file -> Ladybug graph projection."
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.ladybug
  "Ladybug access layer for graph-backed Penpot.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.meta
  "`GraphMeta`: the graph's own account of who built it and from what.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.projection.document
  "Project a Penpot file-data map into Ladybug nodes and structural edges.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.projection.transforms
  "Derived graph links: edges a reader could compute from the projected
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.report
  (:require
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.schema
  "Ladybug DDL facade for the graph-backed Penpot vertical slice.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.schema.contract
  "Deliberate choices in Penpot's graph schema, recorded as data.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.schema.nodes
  "Single source of truth for graph node tables.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.schema.projection
  "Derive Ladybug node column schemas from Penpot Malli sources.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.schema.types
  "Map Malli schemas to Ladybug column types.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.schema.values
  "Shape a Penpot value into the plain data its Ladybug column type wants.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.stats
  (:require
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.graph.sync
  "Incremental Ladybug graph updates from Penpot file-change events."
@@ -15,6 +15,7 @@
  (:require
   [app.common.schema :as sm]
   [app.util.ssrf :as ssrf]
+   [app.worker :as-alias wrk]
   [cuerdas.core :as str]
   [integrant.core :as ig]
   [java-http-clj.core :as http])
@@ -23,6 +24,8 @@
   java.net.URI))

 (def default-max-redirects 5)
+(def default-connect-timeout 30000)
+(def default-request-timeout 30000)

 (defn client?
  [o]
@@ -33,15 +36,17 @@
  :pred client?})

 (defmethod ig/init-key ::client
-  [_ _]
-  (http/build-client {:connect-timeout 30000
+  [_ {:keys [::wrk/executor]}]
+  (http/build-client {:connect-timeout default-connect-timeout
+                      :executor executor
                      :follow-redirects :never}))

 (defn send!
  ([client req] (send! client req {}))
  ([client req {:keys [response-type] :or {response-type :string}}]
   (assert (client? client) "expected valid http client")
-   (http/send req {:client client :as response-type})))
+   (http/send (merge {:timeout default-request-timeout} req)
+              {:client client :as response-type})))

 (defn- resolve-client
  [params]
@@ -60,7 +60,13 @@

 (defmethod handle-error :restriction
  [err request _]
-  (let [{:keys [code] :as data} (ex-data err)]
+  (let [data    (ex-data err)
+        code    (get data :code)
+        explain (ex/explain data)
+        data    (-> data
+                    (dissoc ::sm/explain)
+                    (cond-> explain (assoc :explain explain)))]
+
    (if (= code :method-not-allowed)
      {::yres/status 405
       ::yres/body data}
@@ -36,6 +36,9 @@
 ;; Default age for automatic session renewal
 (def default-renewal-max-age (ct/duration {:hours 6}))

+;; Default absolute maximum session duration
+(def default-cookie-max-age-absolute (ct/duration {:days 30}))
+
 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
 ;; PROTOCOLS
 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
@@ -169,15 +172,19 @@

 (defn- assign-token
  [cfg session]
-  (let [claims {:iss "authentication"
-                :aud "penpot"
-                :sid (:id session)
-                :iat (:modified-at session)
-                :uid (:profile-id session)
-                :sso-provider-id (:sso-provider-id session)
-                :sso-session-id (:sso-session-id session)}
-        header {:kid 1 :ver 1}
-        token  (tokens/generate cfg claims header)]
+  (let [absolute-max-age (cf/get :auth-token-cookie-max-age-absolute default-cookie-max-age-absolute)
+        claims            {:iss "authentication"
+                           :aud "penpot"
+                           :sid (:id session)
+                           :iat (:modified-at session)
+                           :uid (:profile-id session)
+                           :sso-provider-id (:sso-provider-id session)
+                           :sso-session-id (:sso-session-id session)}
+        claims            (if (:created-at session)
+                            (assoc claims :exp (ct/plus (:created-at session) absolute-max-age))
+                            claims)
+        header            {:kid 1 :ver 1}
+        token             (tokens/generate cfg claims header)]
    (assoc session :token token)))

 (defn create-fn
@@ -353,15 +360,23 @@
       or (updated_at is null and
           created_at < ?::timestamptz)")

+(def ^:private
+  sql:delete-expired-v2
+  "DELETE FROM http_session_v2
+    WHERE created_at < ?::timestamptz")
+
 (defn- collect-expired-tasks
  [{:keys [::db/conn ::tasks/max-age]}]
  (let [threshold (ct/minus (ct/now) max-age)
-        result    (-> (db/exec-one! conn [sql:delete-expired threshold threshold])
-                      (db/get-update-count))]
+        result-legacy (-> (db/exec-one! conn [sql:delete-expired threshold threshold])
+                          (db/get-update-count))
+        result-v2     (-> (db/exec-one! conn [sql:delete-expired-v2 threshold])
+                          (db/get-update-count))]
    (l/dbg :task "gc"
           :hint "clean http sessions"
-           :deleted result)
-    result))
+           :deleted-legacy result-legacy
+           :deleted-v2 result-v2)
+    (+ result-legacy result-v2)))

 (defmethod ig/init-key ::tasks/gc
  [_ {:keys [::tasks/max-age] :as cfg}]
@@ -205,7 +205,7 @@
    ::sto/storage (ig/ref ::sto/storage)}

   ::http.client/client
-   {}
+   {::wrk/executor (ig/ref ::wrk/executor)}

   ::session/manager
   {::db/pool (ig/ref ::db/pool)}
@@ -75,10 +75,10 @@
                           {:method method
                            :uri uri
                            :body body
-                            :headers headers}
+                            :headers headers
+                            :timeout timeout}
                           {:response-type :input-stream
-                            :skip-ssrf-check? true
-                            :timeout timeout})
+                            :skip-ssrf-check? true})
            status (:status resp)]
        (when (not (<= 200 status 299))
          (let [body (:body resp)]
@@ -390,18 +390,26 @@

 (def ^:private sql:file-comment-users
  "WITH available_profiles AS (
-     SELECT DISTINCT owner_id AS id
-       FROM comment
-      WHERE thread_id IN (SELECT id FROM comment_thread WHERE file_id=?)
+     SELECT DISTINCT c.owner_id AS id
+     FROM comment c
+     JOIN comment_thread ct
+       ON ct.id = c.thread_id
+    WHERE ct.file_id = ?::uuid
+  ),
+  profile_ids AS (
+    SELECT id FROM available_profiles
+    UNION
+    SELECT ?::uuid
  )
  SELECT p.id,
         p.email,
         p.fullname AS name,
-         p.fullname AS fullname,
+         p.fullname,
         p.photo_id,
         p.is_active
-    FROM profile AS p
-   WHERE p.id IN (SELECT id FROM available_profiles) OR p.id=?")
+    FROM profile p
+    JOIN profile_ids AS x
+      ON x.id = p.id;")

 (defn get-file-comments-users
  [conn file-id profile-id]
@@ -339,6 +339,9 @@

 ;; --- Chunked Upload: Upload a single chunk

+(declare ^:private get-upload-chunk)
+(declare ^:private check-upload-chunk-slot)
+
 (def ^:private schema:upload-chunk
  [:map {:title "upload-chunk"}
   [:session-id ::sm/uuid]
@@ -354,9 +357,31 @@
  {::doc/added "2.17"
   ::sm/params schema:upload-chunk
   ::sm/result schema:upload-chunk-result}
-  [{:keys [::db/pool] :as cfg}
-   {:keys [::rpc/profile-id session-id index content] :as _params}]
-  (let [session (db/get pool :upload-session {:id session-id :profile-id profile-id})]
+  [cfg {:keys [::rpc/profile-id session-id index content]}]
+  (let [session (db/tx-run! cfg check-upload-chunk-slot session-id profile-id index content)]
+    (l/trc :hint "upload-chunk"
+           :session-id session-id
+           :chunk (str index "/" (:total-chunks session))
+           :size (:size content)
+           :path (:path content))
+
+    (let [storage (sto/resolve cfg)
+          data    (sto/content (:path content))]
+      (sto/put-object! storage
+                       {::sto/content      data
+                        ::sto/deduplicate? false
+                        ::sto/touch        true
+                        :content-type      (:mtype content)
+                        :bucket            sto/tempfile-bucket
+                        :upload-id         (str session-id)
+                        :chunk-index       index}))
+
+    {:session-id session-id
+     :index      index}))
+
+(defn- check-upload-chunk-slot
+  [{:keys [::db/conn]} session-id profile-id index content]
+  (let [session (db/get conn :upload-session {:id session-id :profile-id profile-id} {::db/for-update true})]
    (when (or (neg? index) (>= index (:total-chunks session)))
      (ex/raise :type :validation
                :code :invalid-chunk-index
@@ -365,26 +390,23 @@
                :total-chunks (:total-chunks session)
                :index index))

+    (when (> (:size content) (cf/get :upload-max-chunk-size))
+      (ex/raise :type :validation
+                :code :chunk-too-large
+                :hint "chunk size exceeds the maximum allowed"
+                :session-id session-id
+                :index index
+                :size (:size content)
+                :max-size (cf/get :upload-max-chunk-size)))

-    (l/trc :hint "upload-chunk"
-           :session-id session-id
-           :chunk (str index "/" (:total-chunks session))
-           :size (:size content)
-           :path (:path content)))
+    (when (get-upload-chunk conn session-id index)
+      (ex/raise :type :validation
+                :code :duplicate-chunk-index
+                :hint "chunk index already uploaded for this session"
+                :session-id session-id
+                :index index))

-  (let [storage (sto/resolve cfg)
-        data    (sto/content (:path content))]
-    (sto/put-object! storage
-                     {::sto/content      data
-                      ::sto/deduplicate? false
-                      ::sto/touched-at   (ct/in-future {:hours 1})
-                      :content-type      (:mtype content)
-                      :bucket            sto/tempfile-bucket
-                      :upload-id         (str session-id)
-                      :chunk-index       index}))
-
-  {:session-id session-id
-   :index      index})
+    session))

 ;; --- Chunked Upload: shared helpers

@@ -400,6 +422,18 @@
  [conn session-id]
  (db/exec! conn [sql:get-upload-chunks (str session-id)]))

+(def ^:private sql:get-upload-chunk
+  "SELECT id
+     FROM storage_object
+    WHERE (metadata->>'~:upload-id') = ?::text
+      AND (metadata->>'~:chunk-index')::integer = ?
+      AND deleted_at IS NULL
+    LIMIT 1")
+
+(defn- get-upload-chunk
+  [conn session-id index]
+  (db/exec-one! conn [sql:get-upload-chunk (str session-id) index]))
+
 (defn- concat-chunks
  "Reads all chunk storage objects in order and writes them to a single
  temporary file on the local filesystem. Returns a path to that file."
@@ -418,18 +452,21 @@
  conforming to `media.v/schema:upload` with `:filename`, `:path` and
  `:size`.

-  Raises a :validation/:missing-chunks error when the number of stored
-  chunks does not match `:total-chunks` recorded in the session row.
+  Raises a :validation/:missing-chunks error when the stored chunk
+  indices do not form exactly the `0..total-chunks` range recorded in
+  the session row (wrong count, gaps or duplicates).
  Raises :not-found when the session does not belong to `profile-id`.
  Deletes the session row from `upload_session` on success."
  [{:keys [::db/conn] :as cfg} profile-id session-id]
  (let [session (db/get conn :upload-session {:id session-id :profile-id profile-id})
-        chunks  (get-upload-chunks conn session-id)]
+        chunks  (get-upload-chunks conn session-id)
+        indices (sort (map :chunk-index chunks))]

-    (when (not= (count chunks) (:total-chunks session))
+    (when (or (not= (count chunks) (:total-chunks session))
+              (not= indices (range (:total-chunks session))))
      (ex/raise :type :validation
                :code :missing-chunks
-                :hint "number of stored chunks does not match expected total"
+                :hint "stored chunks do not match expected total"
                :session-id session-id
                :expected   (:total-chunks session)
                :found      (count chunks)))
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.rpc.commands.plugins
  (:require
@@ -41,6 +41,7 @@
   [app.rpc.notifications :as notifications]
   [app.storage :as sto]
   [app.util.services :as sv]
+   [app.util.ssrf :as ssrf]
   [app.worker :as wrk]
   [cuerdas.core :as str]))

@@ -960,13 +961,18 @@ RETURNING id, deleted_at;")
 (sv/defmethod ::check-organization-sso
  "Validate an organization SSO configuration by generating a login redirect URL.
  Nitrate calls this while configuring SSO to verify client credentials and OIDC
-  discovery before saving the settings."
+  discovery before saving the settings. The issuer URL is nitrate-supplied
+  (customer-configured), so it is checked against the SSRF blocklist before
+  any outbound request is attempted."
  {::doc/added "2.18"
   ::sm/params cto/schema:nitrate-sso
   ::sm/result schema:check-organization-sso-result
   ::rpc/auth false}
  [cfg params]
-  {:valid (oidc/is-organization-sso-config-valid? cfg params)})
+  (let [issuer (oidc/organization-sso-discovery-uri params)]
+    {:valid (boolean (and issuer
+                          (ssrf/safe-url? issuer)
+                          (oidc/is-organization-sso-config-valid? cfg params)))}))

 ;; ---- API: notify-organization-sso-change
 (sv/defmethod ::notify-organization-sso-change
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.storage.pending-gc
  "A maintenance task that reclaims storage objects created in 'pending'
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns app.tasks.demo-purge
  "Task handler for delayed demo profile deletion. Submitted at demo
@@ -17,7 +17,6 @@
   [app.http.client :as http]
   [app.main :as-alias main]
   [app.setup :as-alias setup]
-   [app.util.blob :as blob]
   [app.util.json :as json]
   [integrant.core :as ig]
   [promesa.exec :as px]))
@@ -248,20 +247,16 @@
    :props      (or (some-> props db/decode-transit-pgobject) {})
    :context    (or (some-> context db/decode-transit-pgobject) {})}))

-(defn- encode-batch
-  "Encode a sequence of event maps into a fressian+zstd base64 string
-  suitable for JSON transport."
-  ^String [events]
-  (blob/encode-str events {:version 4}))
-
 (defn send-event-batch
  "Send a single batch of events to the telemetry endpoint. Returns
-  true on success."
+  true on success. The events are sent as a plain vector of event
+  maps; the JSON encoder handles UUID and temporal types natively and
+  the receiver coerces them back to proper types."
  [{:keys [::setup/props] :as cfg} batch]
  (let [payload {:type :telemetry-events
                 :version (:full cf/version)
                 :instance-id (:instance-id props)
-                 :events (encode-batch batch)}
+                 :events (vec batch)}
        request {:method  :post
                 :uri     (cf/get :telemetry-uri)
                 :headers {"content-type" "application/json"}
@@ -0,0 +1,76 @@
+;; This Source Code Form is subject to the terms of the Mozilla Public
+;; License, v. 2.0. If a copy of the MPL was not distributed with this
+;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
+;;
+;; Copyright (c) KALEIDOS SUBSIDIARY SL
+
+(ns backend-tests.auth-ldap-test
+  (:require
+   [app.auth.ldap :as ldap-auth]
+   [clj-ldap.client :as ldap]
+   [clojure.test :as t]))
+
+;; --- search-user: filter must be escaped (RED: currently not escaped)
+
+(t/deftest search-user-escapes-email-in-filter
+  (t/testing "wildcard * is escaped before building LDAP filter"
+    (let [captured-query (atom nil)
+          fake-search    (fn [_conn _base-dn params]
+                           (reset! captured-query (:filter params))
+                           [])]
+      (with-redefs [ldap/search fake-search]
+        (#'ldap-auth/search-user {:query "(mail=:username)" :sizelimit 1
+                                  :attrs-username "uid" :attrs-email "mail"
+                                  :attrs-fullname "cn"}
+                                 "fry*@planetexpress.com"))
+      ;; After fix: * should be escaped as \2a
+      (t/is (= "(mail=fry\\2a@planetexpress.com)" @captured-query)
+            "filter must have * escaped per RFC 4515"))))
+
+;; --- retrieve-user: email must come from directory, not client (RED)
+
+(t/deftest retrieve-user-uses-directory-email
+  (t/testing "returned email is from LDAP directory, not client input"
+    (let [fake-search (fn [_conn _base-dn _params]
+                        [{:dn "cn=fry,ou=people,dc=planetexpress,dc=com"
+                          :mail "fry@planetexpress.com"
+                          :cn "Philip J. Fry"
+                          :uid "fry"}])
+          fake-bind?  (fn [_conn _dn _password] true)]
+      (with-redefs [ldap/search fake-search
+                    ldap/bind?  fake-bind?]
+        (let [cfg     {:query "(mail=:username)" :sizelimit 1
+                       :attrs-username "uid" :attrs-email "mail"
+                       :attrs-fullname "cn"}
+              result  (#'ldap-auth/retrieve-user cfg {:email "fry*@planetexpress.com" :password "fry"})]
+          ;; After fix: email should be from directory (fry@planetexpress.com)
+          ;; BUG: email is client input (fry*@planetexpress.com)
+          (t/is (= "fry@planetexpress.com" (:email result))
+                "email must come from LDAP directory attribute, not client input"))))))
+
+;; --- authenticate: full flow with directory email (RED)
+
+(t/deftest authenticate-returns-directory-email
+  (t/testing "authenticate returns directory email for profile"
+    (let [fake-search (fn [_conn _base-dn _params]
+                        [{:dn "cn=amy,ou=people,dc=planetexpress,dc=com"
+                          :mail "amy@planetexpress.com"
+                          :cn "Amy Wong"
+                          :uid "amy"}])
+          fake-bind?  (fn [_conn _dn _password] true)]
+      (with-redefs [ldap/search fake-search
+                    ldap/bind?  fake-bind?
+                    ldap/connect (fn [_cfg] (reify java.lang.AutoCloseable (close [_] nil)))]
+        (let [cfg    {:query "(mail=:username)" :sizelimit 1
+                      :attrs-username "uid" :attrs-email "mail"
+                      :attrs-fullname "cn"
+                      :bind-dn "cn=admin,dc=planetexpress,dc=com"
+                      :bind-password "GoodNewsEveryone"
+                      :host "localhost" :port 10389
+                      :ssl false :tls false
+                      :base-dn "ou=people,dc=planetexpress,dc=com"}
+              result (ldap-auth/authenticate cfg {:email "*@planetexpress.com" :password "amy"})]
+          ;; After fix: email should be amy@planetexpress.com (directory)
+          ;; BUG: email is *@planetexpress.com (client)
+          (t/is (= "amy@planetexpress.com" (:email result))
+                "authenticate must return directory email, not client-supplied wildcard"))))))
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns backend-tests.demo-test
  (:require
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns backend-tests.graph-binder-gate-test
  "Binder gate for the incremental-sync statement templates.
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns backend-tests.graph-sync-parity-test
  "Cold projection and incremental sync are two implementations of one mapping,
@@ -0,0 +1,30 @@
+;; This Source Code Form is subject to the terms of the Mozilla Public
+;; License, v. 2.0. If a copy of the MPL was not distributed with this
+;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
+;;
+;; Copyright (c) KALEIDOS SUBSIDIARY SL
+
+(ns backend-tests.http-client-test
+  (:require
+   [app.http.client :as http]
+   [clojure.test :as t]
+   [java-http-clj.core :as jhttp]
+   [mockery.core :refer [with-mocks]]))
+
+(t/deftest send-injects-default-timeout-when-absent
+  (with-mocks [mock {:target 'java-http-clj.core/send
+                     :return {:status 200 :body ""}}]
+    (let [client (jhttp/build-client {})]
+      (http/send! client {:method :get :uri "https://example.com/"})
+      (let [[req _opts] (:call-args @mock)]
+        (t/is (= http/default-request-timeout (:timeout req)))))))
+
+(t/deftest send-preserves-caller-supplied-timeout
+  (with-mocks [mock {:target 'java-http-clj.core/send
+                     :return {:status 200 :body ""}}]
+    (let [client (jhttp/build-client {})]
+      (http/send! client {:method :get
+                          :uri "https://example.com/"
+                          :timeout 5000})
+      (let [[req _opts] (:call-args @mock)]
+        (t/is (= 5000 (:timeout req)))))))
@@ -277,6 +277,70 @@
    (t/is (= (:id session) (:sid claims)))
    (t/is (= (:id profile) (:uid claims)))))

+(t/deftest session-token-contains-exp-claim
+  (let [cfg     th/*system*
+        manager (session/inmemory-manager)
+        profile (th/create-profile* 1)
+        session (->> (session/create-session manager {:profile-id (:id profile)
+                                                      :user-agent "user agent"})
+                     (#'session/assign-token cfg))
+        claims  (tokens/decode cfg (:token session))
+        exp     (:exp claims)]
+    (t/is (some? exp) "session token should contain :exp claim")
+    (t/is (ct/inst? exp) "exp should be an instant")))
+
+(t/deftest session-token-exp-based-on-created-at
+  (let [cfg              th/*system*
+        manager          (session/inmemory-manager)
+        profile          (th/create-profile* 1)
+        session          (->> (session/create-session manager {:profile-id (:id profile)
+                                                               :user-agent "user agent"})
+                              (#'session/assign-token cfg))
+        claims           (tokens/decode cfg (:token session))
+        expected-exp     (ct/plus (:created-at session) (ct/duration {:days 30}))]
+    (t/is (some? (:exp claims)) "session token should contain :exp claim")
+    (t/is (= (inst-ms (:exp claims))
+             (inst-ms expected-exp))
+          "exp should equal created-at + 30 days")))
+
+(t/deftest session-token-past-exp-is-rejected
+  (let [cfg     th/*system*
+        manager (session/inmemory-manager)
+        profile (th/create-profile* 1)
+        session (->> (session/create-session manager {:profile-id (:id profile)
+                                                      :user-agent "user agent"})
+                     (#'session/assign-token cfg))
+        claims  (tokens/decode cfg (:token session))
+        ;; Manually create a token with exp in the past
+        past-claims (assoc claims :exp (ct/minus (ct/now) (ct/duration {:days 1})))
+        header     {:kid 1 :ver 1}
+        past-token (tokens/generate cfg past-claims header)]
+    (t/is (nil? (session/decode-token cfg past-token))
+          "token with exp in the past should be rejected")))
+
+(t/deftest session-renewal-preserves-original-exp
+  (let [cfg      th/*system*
+        manager  (session/inmemory-manager)
+        profile  (th/create-profile* 1)
+        handler  (-> (fn [req] req)
+                     (#'session/wrap-authz  {::session/manager manager})
+                     (#'mw/wrap-auth {:bearer (partial session/decode-token cfg)
+                                      :cookie (partial session/decode-token cfg)}))
+        session  (->> (session/create-session manager {:profile-id (:id profile)
+                                                       :user-agent "user agent"})
+                      (#'session/assign-token cfg))
+        original-exp (:exp (tokens/decode cfg (:token session)))
+        ;; Force renewal by setting modified-at to 7 hours ago
+        old-session  (assoc session :modified-at (ct/minus (ct/now) (ct/duration {:hours 7})))
+        response    (handler (make-dummy-request {:cookies {"auth-token" (:token old-session)}}))
+        {:keys [token claims]} (get response ::http/auth-data)
+        new-exp     (:exp claims)]
+    (t/is (some? original-exp) "original token should have :exp")
+    (t/is (some? new-exp) "renewed token should have :exp")
+    (t/is (= (inst-ms original-exp)
+             (inst-ms new-exp))
+          "renewed token should preserve original :exp, not extend it")))
+
 (t/deftest parse-request-illegal-argument-exception
  ;; clojure.data.json raises IllegalArgumentException (case
  ;; fall-through) on several kinds of malformed input. The
@@ -8,6 +8,7 @@
  (:require
   [app.common.exceptions :as ex]
   [app.config :as cf]
+   [app.http.client :as http]
   [app.media.remote :as media.remote]
   [app.setup :as-alias setup]
   [app.util.json :as json]
@@ -500,6 +501,22 @@
                     :headers {}})]
          (t/is (= 200 (:status resp))))))))

+(t/deftest service-request-puts-configured-timeout-in-request
+  (t/testing "service-request puts media-processing-service-timeout on the http request"
+    (let [captured (atom nil)]
+      (with-redefs [cf/get (th/config-get-mock config-mock)
+                    http/req (fn [_client request _opts]
+                               (reset! captured request)
+                               {:status 200
+                                :body (json-stream {:width 100 :height 100})})]
+        (media.remote/service-request
+         (mk-system)
+         {:method :post
+          :uri "http://localhost:6065/api/image/info"
+          :body nil
+          :headers {}})
+        (t/is (= 5000 (:timeout @captured)))))))
+
 ;; ---------------------------------------------------------------------------
 ;; Shared key
 ;; ---------------------------------------------------------------------------
@@ -2,7 +2,7 @@
 ;; License, v. 2.0. If a copy of the MPL was not distributed with this
 ;; file, You can obtain one at http://mozilla.org/MPL/2.0/.
 ;;
-;; Copyright (c) KALEIDOS INC Sucursal en España SL
+;; Copyright (c) KALEIDOS SUBSIDIARY SL

 (ns backend-tests.passwords-test
  (:require