Files
LocalAI/pkg/mcp/localaitools/client.go
T
Richard Palethorpe d10374f849 feat(router): make KNN a first-class classifier with a persisted, curated corpus (#10652)
* feat(router): make KNN a first-class classifier with a persisted, curated corpus

Add `classifier: knn` — similarity-weighted voting over labelled
example prompts. Unlike score/colbert it needs no classifier model:
label knowledge lives in a corpus seeded and curated through the
admin API, so routing decisions are deterministic, auditable, and
grounded in graded experience rather than a model's opinion.

Epistemic gate: corpus entries below knn.similarity_threshold cannot
vote; when none clears it the classifier activates no labels and the
router uses the fallback — a prompt unlike all labelled experience is
treated as undecidable, not guessed. Decisions record
nearest_similarity (also on fallback rows) so admins can see how far
the nearest labelled experience was; the Routing tab explains
out-of-corpus fallbacks and shows per-label corpus counts.

Persistence: one JSONL file per router under
<data path>/router-corpus (text, labels, vector, embedder
fingerprint). The file is the source of truth; the local-store index
is rebuilt from it at classifier build time and stays a pure
in-memory index. Entries recorded under a different embedding model
re-embed on load. Also corrects the docs' false claim that
local-store collections persist — the embedding cache never survived
restarts (and still doesn't); the corpus does.

Corpus input is API-only by design (entries may contain example user
content): POST /api/router/{name}/corpus seeds (labels validated
against declared policies, embedded server-side, indexed
immediately), GET .../corpus/stats inspects — label counts only,
entry texts are never returned by any surface — DELETE .../corpus
wipes. Admin-gated like the sibling router endpoints, and exposed as
MCP tools (seed_router_corpus / get_router_corpus_stats /
clear_router_corpus) in both the httpapi and inproc clients with
coverage-test route mappings.

Plumbing: VectorStore gains SearchK (top-K was hardcoded to 1);
local-store gets InsertBatch/Delete as optional fast paths;
RouterConfig gains a knn block (embedding_model, k,
similarity_threshold, vote_threshold, store_name) with meta-registry
fields; the classifier dropdown now offers knn and the
previously-missing colbert; embedding_cache is ignored (with a
warning) for knn — it IS an embedding-KNN lookup; the stale
/api/instructions intelligent-routing entry is rewritten (it
described a classifier that no longer exists); swagger regenerated.

Tests: KNN vote/gate specs with hand-computed vote shares, corpus
manager suite (restart reload without re-embedding, fingerprint
re-embed, dedupe, hostile store names), middleware specs (corpus
routing, gate fallback, config validation, cache-wrap refusal),
corpus endpoint specs pinning the texts-never-returned contract, MCP
catalog + route-mapping gates, and a Playwright spec for corpus
stats and the out-of-corpus decision detail.

Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* feat(router): name consulted corpus neighbours in knn decisions

Every knn decision (decision log rows and the /api/router/decide
response) now carries neighbors: the K retrieved corpus entries by
descending similarity - including ones below the epistemic gate, which
is what makes fallback decisions diagnosable - each as {id, similarity,
labels}. The id is the entry's content hash (first 8 bytes of the
SHA-256 of its text, hex): stable across reseeds and re-embeds, and
text-free, so an external platform that seeded the corpus can recompute
text->id on its own copy and bucket decisions by corpus region (per-
region reliability accounting) without corpus text ever leaving the
server. A corrupt index payload surfaces as an id-less neighbour at a
real similarity instead of disappearing.

Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* refactor(router): deduplicate knn plumbing and cut corpus hot-path waste

Post-review cleanup of the knn-first-class-router branch; no behaviour
changes on the API surface.

Reuse/altitude:
- RouterKNNConfig.ResolvedStoreName is now the single source of the
  router-corpus-<name> default (was hand-derived in four files).
- corpus.ResolveKNNRouter + corpus.Seed carry the shared model
  resolution and seed validation; the REST endpoints and the assistant
  MCP client are thin transport adapters over them, with sentinel
  errors mapped to HTTP statuses at the echo boundary.
- middleware.NewClassifierDeps assembles the classifier dependency set
  once for all five entry points (OpenAI, Anthropic, realtime, decide,
  corpus) instead of five hand-copied literals.
- router.AllClassifiers feeds both the status endpoint and the
  unknown-classifier error, ending the classifier-list drift.
- Per-classifier requirements moved out of validateRouterPolicies into
  their buildClassifier arms; the knn arm owns its embedding_cache
  opt-out instead of a name-check in the shared wrap tail.
- adminOnly replaces four inline copies of the admin gate in the
  middleware routes.
- localVectorStore.Search delegates to SearchK (identical traces).

Efficiency:
- Manager.Add embeds outside the manager mutex and appends to the
  JSONL file (O(new) instead of O(corpus) rewrite); a torn tail from a
  crash mid-append is tolerated on read and repaired on next write.
- Stats memoises per store keyed on the file's stat fingerprint and no
  longer takes the manager mutex, so the 5s status poll stops parsing
  vector-laden JSONL and stops blocking behind seeds.
- KNN Classify decodes each neighbour payload once (was twice) and
  builds refs and votes in a single pass with one fallback return.
- Corpus file writes fsync before rename/close.
- The corpus manager is built eagerly in newApplication (sync.Once
  dropped); test helper dead branch removed.

Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* feat(router): bind knn corpus vectors to an embedder fingerprint and fail closed on mismatch

Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* chore(mcp): align corpus tool prompts and the mutating-tool safety list

Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* feat(proto,backend): report embedding shape from the llama-cpp backend

Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* feat(embeddings): Go-side pooling — mean/last/decayed_mean with half-life

Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* feat(embeddings): accept chat messages[] and per-request pooling on /v1/embeddings

Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* chore(middleware): name the failing fields when post-merge validation 400s

An intermittent post-merge validation failure surfaced as an opaque 400
during integration (pooling scheme mismatch that no client had sent).
Log the model, the request's pooling override, and the merged config's
pooling fields at the failure point so the next occurrence identifies
whether the request or the stored config carried the bad value.

Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* fix(embeddings): scheme override must not inherit the config's half-life

A model config defaulting to decayed_mean pooling carries
pooling_half_life_tokens; a request overriding the scheme to mean/last
without its own half-life inherited that value, and post-merge
validation rejected the pair the server itself had assembled. Zero the
inherited half-life when the overridden scheme is not decayed_mean; a
request that explicitly pairs a half-life with a non-decayed scheme
still 400s.

Assisted-by: Claude:claude-fable-5 [Claude Code]
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* fix embedding pooling validation and router bounds

Declare backend embedding layouts and reject incompatible pooling modes. Reset local-store dimensions after a full clear, validate KNN thresholds, and add real backend and store integration coverage.

Assisted-by: Codex:gpt-5
Signed-off-by: Richard Palethorpe <io@richiejp.com>

* ci: run local-store integration tests

Build and install the local-store backend in the Linux test job, then run the existing store integration suite so new specs are discovered automatically.

Assisted-by: Codex:gpt-5
Signed-off-by: Richard Palethorpe <io@richiejp.com>

---------

Signed-off-by: Richard Palethorpe <io@richiejp.com>
2026-08-18 09:37:43 +02:00

134 lines
6.6 KiB
Go

package localaitools
import (
"context"
"github.com/mudler/LocalAI/core/config"
"github.com/mudler/LocalAI/core/gallery"
"github.com/mudler/LocalAI/core/schema"
"github.com/mudler/LocalAI/core/services/modeladmin"
"github.com/mudler/LocalAI/pkg/vram"
)
// LocalAIClient is the surface tools depend on. It has two implementations:
//
// - inproc.Client (in-process; calls LocalAI services directly)
// - httpapi.Client (out-of-process; calls the LocalAI REST API)
//
// Tool handlers and the embedded skill prompts are agnostic to which
// implementation backs the client.
//
// Where the same shape already exists elsewhere in the codebase
// (config.Gallery, gallery.Metadata, schema.KnownBackend, vram.EstimateResult,
// modeladmin.Action/Capability) we surface it directly rather than maintain
// a parallel DTO — keeping the LLM-visible wire format aligned with the
// rest of LocalAI by construction.
type LocalAIClient interface {
// ---- Models / gallery (read) ----
GallerySearch(ctx context.Context, q GallerySearchQuery) ([]gallery.Metadata, error)
ListInstalledModels(ctx context.Context, capability Capability) ([]InstalledModel, error)
ListGalleries(ctx context.Context) ([]config.Gallery, error)
GetJobStatus(ctx context.Context, jobID string) (*JobStatus, error)
GetModelConfig(ctx context.Context, name string) (*ModelConfigView, error)
// ---- Models / gallery (write) ----
InstallModel(ctx context.Context, req InstallModelRequest) (jobID string, err error)
DeleteModel(ctx context.Context, name string) error
EditModelConfig(ctx context.Context, name string, patch map[string]any) error
ReloadModels(ctx context.Context) error
// LoadModel pre-loads a model into memory by name (the inverse of shutting
// it down). For a realtime pipeline model every configured sub-model is
// loaded; it returns the model names that became resident.
LoadModel(ctx context.Context, model string) ([]string, error)
ImportModelURI(ctx context.Context, req ImportModelURIRequest) (*ImportModelURIResponse, error)
// ---- Model aliases ----
// SetAlias creates the alias `name` pointing at `target`, or swaps an
// existing alias's target. The server validates that `target` is an
// existing, non-alias, enabled model. Deletion reuses DeleteModel.
SetAlias(ctx context.Context, name, target string) error
// ListAliases returns every configured alias and its target.
ListAliases(ctx context.Context) ([]AliasInfo, error)
// ---- Backends ----
// ListBackends returns installed backends. The shape stays a thin
// localaitools.Backend rather than gallery.SystemBackend because the
// latter carries filesystem paths (RunFile, Metadata) the LLM
// shouldn't see.
ListBackends(ctx context.Context) ([]Backend, error)
// ListKnownBackends returns the same shape as REST /backends/known.
ListKnownBackends(ctx context.Context) ([]schema.KnownBackend, error)
InstallBackend(ctx context.Context, req InstallBackendRequest) (jobID string, err error)
UpgradeBackend(ctx context.Context, name string) (jobID string, err error)
// ---- System ----
SystemInfo(ctx context.Context) (*SystemInfo, error)
ListNodes(ctx context.Context) ([]Node, error)
ListScheduling(ctx context.Context) ([]ModelSchedulingConfig, error)
GetScheduling(ctx context.Context, modelName string) (*ModelSchedulingConfig, error)
SetScheduling(ctx context.Context, req SetSchedulingRequest) (*ModelSchedulingConfig, error)
DeleteScheduling(ctx context.Context, modelName string) error
// SetNodeVRAMBudget sets (or, with an empty budget, clears) a federated
// node's VRAM allocation cap as a sticky admin override. Only meaningful
// in distributed mode; single-process clients report it as unavailable.
SetNodeVRAMBudget(ctx context.Context, nodeID, budget string) error
VRAMEstimate(ctx context.Context, req VRAMEstimateRequest) (*vram.EstimateResult, error)
// ---- State ----
// ToggleModelState accepts modeladmin.ActionEnable / ActionDisable.
ToggleModelState(ctx context.Context, name string, action modeladmin.Action) error
// ToggleModelPinned accepts modeladmin.ActionPin / ActionUnpin.
ToggleModelPinned(ctx context.Context, name string, action modeladmin.Action) error
// ---- Branding / whitelabeling ----
// GetBranding returns the configured instance branding (name, tagline,
// asset URLs).
GetBranding(ctx context.Context) (*Branding, error)
// SetBranding updates the text branding fields. Asset uploads are not
// exposed over MCP — admins use the Settings UI for binary files.
SetBranding(ctx context.Context, req SetBrandingRequest) (*Branding, error)
// ---- Voice profile library ----
ListVoiceProfiles(ctx context.Context) ([]VoiceProfile, error)
CreateVoiceProfile(ctx context.Context, req CreateVoiceProfileRequest) (*VoiceProfile, error)
DeleteVoiceProfile(ctx context.Context, id string) error
// ---- Usage / billing ----
// GetUsageStats returns aggregated token usage. In single-user
// no-auth mode this reports the synthetic local user's usage. The
// implementation enforces "admin required to query other users".
GetUsageStats(ctx context.Context, q UsageStatsQuery) (*UsageStats, error)
// ---- PII filter ----
// GetPIIEvents returns recent redaction events. Implementation
// enforces "admin required" when auth is on. The regex pattern tools
// were removed — detection policy lives on each detector model's
// pii_detection block, managed via the model-config tools.
GetPIIEvents(ctx context.Context, q PIIEventsQuery) ([]PIIEvent, error)
// ---- Middleware admin ----
// GetMiddlewareStatus returns the aggregated state surfaced on the
// /app/middleware page: active PII patterns, per-model resolved
// enabled state, recent event count, router placeholder.
GetMiddlewareStatus(ctx context.Context) (*MiddlewareStatus, error)
// ---- Router (intelligent routing) ----
// GetRouterDecisions returns recent routing decisions for the
// /app/middleware Routing tab and for agent-driven introspection.
// Admin-required when auth is on.
GetRouterDecisions(ctx context.Context, q RouterDecisionsQuery) ([]RouterDecision, error)
// GetRouterCorpusStats reports a knn router's corpus size and
// per-label counts — counts only, texts are never exposed.
GetRouterCorpusStats(ctx context.Context, routerModel string) (*RouterCorpusStats, error)
// SeedRouterCorpus adds labelled exemplars to a knn router's
// corpus (embedded server-side, persisted, indexed immediately).
SeedRouterCorpus(ctx context.Context, req RouterCorpusSeedRequest) (*RouterCorpusSeedResult, error)
// ClearRouterCorpus wipes a knn router's corpus — file and live
// index.
ClearRouterCorpus(ctx context.Context, routerModel string) (*RouterCorpusClearResult, error)
}