- Introduced `mock-bin-server.ts` for handling mock API requests and responses.
- Added `oauth2-server.ts` to simulate OAuth2 authorization flows with in-memory token management.
- Implemented `socket-server.ts` for WebSocket communication with authentication.
- Created `ws-echo-server.ts` for echoing WebSocket messages.
- Updated `upgrade-fixture.ts` to use `import.meta.dirname` for path resolution.
- Modified `package.json` to switch script files from `.js` to `.ts` for TypeScript compatibility.
- Adjusted Playwright configuration to reference TypeScript files for mock servers.
- Enhanced cloud sync mock to reflect changes in file extensions.
- Updated TypeScript configuration to allow importing TypeScript extensions.
- bind mock-api and echo-server to 127.0.0.1
- remove per-test dataPath after use
- throw when setCodeMirrorValue does not hold after retries
- drop continue-on-error from the build job
- shorten cookie waitForPersisted docstring
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
The pfctl rule that blocks QUIC (UDP 443) was only applied before the
Inso CLI stapling step. The app packaging step also staples through
https://api.apple-cloudkit.com and hits the same issue, so run the
workaround before "Package app (MacOS only)" instead.
Ref: actions/runner-images#14924
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
* ci(ksai): start a KSAI run only for people who can write to the repository
A comment or review now starts the KSAI run job only when its author
can write to this repository. A new authorize job reads the author's
permission with the workflow token, and the run job needs it. A
dispatched run starts as before.
Signed-off-by: Bart Smykla <bartek@smykla.com>
* ci(ksai): fail the authorize step when the permission read fails
A failed permission read used to answer writes=false with a green
job, so a writer's comment started nothing and the log said nothing.
The step now fails on a read error, the run job still does not start,
and the log names the error. An author who cannot write still reads
as writes=false.
Signed-off-by: Bart Smykla <bartek@smykla.com>
---------
Signed-off-by: Bart Smykla <bartek@smykla.com>
A public repository cannot call a reusable workflow in the internal
KSAI repository, so the setup workflow fails before any job starts
with "workflow was not found". Call the published snapshot instead,
pinned to the current release and following its consumer template,
with every command enabled and a review on each opened pull request.
The snapshot names no model gateway of its own, so the call passes the
organization's gateway variable. It runs on GitHub-hosted runners,
because the Kong runner pool serves no public repository, and with no
GitHub App at all: push_as_app is false and no key is passed, so every
read and write goes through GITHUB_TOKEN.
Signed-off-by: Bart Smykla <bartek@smykla.com>
Two independent papercuts in the circular reference check, both hit while
investigating a failure on another PR.
`npm run check-cycle-references` could not run on Windows at all. It spawned
the extensionless `node_modules/.bin/depcruise` shim through `execFileSync`,
which Windows cannot launch; the `.cmd` sibling does not help either, since
Node >=18 refuses to spawn `.cmd` without a shell (CVE-2024-27980). Resolve
dependency-cruiser's own entry point and run it with `process.execPath`
instead — no shell, works everywhere.
The check also fails on baseline drift (cycles recorded in the baseline that
no longer exist), which is correct: the baseline is a ratchet, and leaving a
fixed cycle in it would silently keep permitting its reintroduction. But CI
labelled every non-success outcome "New circular references detected", so a PR
that *removed* two cycles was reported as having added some. That cost real
debugging time.
Give the script distinct exit codes — 1 for a new cycle, 2 for drift alone —
and have the workflow capture and map them to separate messages. The failure
behaviour is unchanged; only the wording is now accurate. Renamed the final
step to match what it actually gates on.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
## Summary
- Replaces the madge-based `check-cycle-references` with a `dependency-cruiser`-based check, cruised once per npm workspace package.
- Keeps the checker, dependency-cruiser config, and known-cycle baseline together under `scripts/circular-references/`.
- Uses a committed baseline to grandfather existing cycles while failing on new cycles and stale baseline entries.
- Cleans the existing `insomnia-inso` type cycle by using `insomnia-data`'s `AllTypes` instead of importing the database container type.
- CI explicitly uses Bash for correct `tee`/pipeline failure propagation, safely handles skipped checks, and posts a PR report.
- Dropped `madge`/`.madgerc`, added `dependency-cruiser`.
actions/checkout steps across CI workflows leaves GITHUB_TOKEN persisted in git config after checkout. None of these steps perform a git push using that persisted credential, disable credential persistence on the actions.
* fix(release): resolve previous tag by semver precedence, not publish order
GitHub's auto-detected previous tag for release notes ignores per-line
release branches, so out-of-order alpha/beta/hotfix tags can pull PRs
from an unrelated release into the notes. Compute the previous core@ tag
via real semver precedence instead, excluding a stable release's own
prereleases so it diffs against the prior stable release.
* fix(release): let ncipollo/release-action generate notes directly
Pass generateReleaseNotesPreviousTag to the release action instead of
calling GitHub's generate-notes API manually via curl - the action
already wraps the same endpoint with identical previous_tag_name
semantics, so this drops the redundant step.
* fix(ci): skip circular refs PR comment when count is unchanged
Avoids noisy comment creation/updates on the circular references check
when the PR introduces no change vs the base branch.
* fix(ci): delete stale circular refs comment when count returns to no-change
Prevents a stale "warning" comment lingering on a PR after circular
references introduced earlier are fixed and the count matches base again.
* chore: emit junit reports per e2e tests shard
* chore: include retries in junit reporter
* chore: test some stuff
* chore: test again
* chore: revert test
* chore: bump easily bumpable actions off of Node 20 runtime
* chore: bump remaining actions or replace with inline logic where Node 24 versions are unavailable
* refactor: remove variable interpolation in run blocks as flagged by Semgrep
* Add vault-crypto/mime utilities and remove heavyweight third-party imports
- Add AES-GCM vault-crypto utility with tests (replaces node-forge usage)
- Add common/mime.ts to replace mime-types package dependency
- Replace tough-cookie import in response-cookies-viewer with inline parser
- Replace @grpc/grpc-js status import in grpc-status-tag with inline constant
- Replace electron.ipcRenderer in auth.clear-vault-key with showToast()
- Remove unused analytics call from window-utils
* Fix impure Date.now() key on CodeEditor; use setValue via ref instead
Replace key={Date.now()} with a useEffect that calls editorRef.current.setValue(snippet)
whenever snippet changes, keeping the editor mounted. Also apply prettier fixes from quick-check.
* fix: address Copilot review comments on PR #9992
- vault-crypto: replace forge-in-renderer with IPC bridge (main process
retains forge; renderer calls window.main.vault.{en,de}cryptSecretValue)
- mime.ts: expand lookup table to 48 entries (webp, wasm, mp4, docx, xlsx,
fonts, audio/video, etc.) and fix remaining mime-types import in send route
- response-viewer: move charset alias map to module level; normalise iconv-lite
alias names (utf8, latin1, win1252, …) to WHATWG labels for TextDecoder
- auth.clear-vault-key: fix typo "all you local" → "all your local"
* fix: sort imports in send route
* feat: disable nodeIntegration in renderer mainWindow, remove import check tooling
- Set nodeIntegration:false and contextIsolation:true on mainWindow webPreferences
(hidden window keeps nodeIntegration:true for user script execution)
- Split script-security-rules.ts out of script-security-policy.ts so the renderer
can import display-only constants without pulling in require-interceptor
- Add templating/renderer-safe.ts with Node-free render/reload/getTagDefinitions;
update all renderer callers to import from it instead of templating/index
- Split insomnia-testing generate.ts: move generateToFile to generate-to-file.ts
so generate() has no Node imports; expose generateToFile from new entry point
- Move runTests execution to main process via IPC (run-tests channel) so the
renderer routes no longer import the Mocha-backed test runner directly
- Delete vite-plugin-electron-node-require.ts, check-renderer-node-imports.ts,
renderer-node-import-baseline.json and all related scripts/plugins now that the
renderer bundle is free of Node built-in imports
* fix: sort imports, use static TestResults type, remove unused analytics import
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* remove mime stuff
* remove ci step
* update plan
* insomnia testing adapter
* use export method
* trick react router ssr
* add renderer errors
* globalThis
* improve error
* move plugin types
* ipc validate proto
* fix import
* plugin types
* polyfill events for jshint
* restore node require plugin
* vault adapter
* add crypto bridges
* tough-cookie ipc
* util stub
* split cookie into network adapter
* assert
* fix plugin index import
* serialise cookie
* decouple renderer from scripting
* Fix rebase conflicts and import path issues
- Fix incorrect ~/insomnia-data imports (should be insomnia-data package)
- Remove non-existent mime utility imports and provide simple fallback
- Remove incorrect analytics call from main process
- Remove unused imports (Settings, Cookie)
- Fix Response type annotation for getResponseBodyBuffer
* lint
* fix tests
* fix: use dynamic import for crypt in session.ts for main process compatibility
The session.ts module is used in both renderer and main process contexts (via sync.invoke IPC handlers). When running in the main process, window.main is undefined, causing TypeError when trying to access window.main.crypt.decryptAES().
Changes:
- Use dynamic import of crypt module (only loaded in main process context)
- In renderer: window.main.crypt is always available so dynamic import never executes
- In main process: dynamic import loads crypt with node:crypto support
- Protect loginStateChange() calls with window existence checks
This avoids bundling node:crypto in the Vite renderer build while still supporting both execution contexts.
Fixes E2E test failures in sync operations (remoteBackendProjects, _assertSession, etc.) caused by disabled nodeIntegration.
* fix: add aria-label to template tag preview and browser-safe encoding fallback
- Add aria-label="Live Preview" to textarea in TagEditor for better Playwright accessibility
- Add atob() fallback for decodeEncoding in browser contexts where Buffer isn't available
- Fixes smoke test element discovery for template tag preview modal
* fix: add explicit waits for element stability in environment test
Add toBeEnabled() and toBeVisible() waits before clicking elements in the
'kv pair environment can be updated' test. This prevents timeout errors from
unstable/flickering elements during modal interactions, especially in
high-concurrency shard execution.
* remove unused
* add autocomplete generate test step
* fix lint
* put analytics back in
* combine vault and crypto adapter
* remove unused export split
* remove insomnia-testing cruft, addressing feedback
* fix: use direct imports in insomnia-inso after insomnia-testing index.ts removal
* fix circular reference
* fix: move createElectronNotifier to main process to fix SSR error
Move the electron-dependent createElectronNotifier function from repo-file-watcher.ts
to git-service.ts to prevent electron imports from being evaluated in the renderer/SSR context.
* export har ipc bridge
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
