mirror of
https://github.com/mudler/LocalAI.git
synced 2026-06-14 11:49:33 -04:00
* fix(router): score classifier production-readiness Conversation trimming runs through the classifier model's chat template and trims by exact token count, sized to the model's n_batch which is now scaled to context so long probes can't crash the backend. Missing chat_message templates are a hard error at router build time. Router- facing factories (Embedder/Scorer/Reranker/TokenCounter) re-resolve ModelConfig per call so a model installed post-startup doesn't bind a stub Backend="" config and silently fall into the loader's auto- iterate path. New 'vector_store' backend trace recorded inside localVectorStore on every Search/Insert — including the backend-load-failure path that previously vanished into an xlog.Warn — with outcome tagging (hit/miss/empty_store/backend_load_error/find_error/insert_error/ok). Companion cleanup drops misleading similarity:0 and input_tokens_count:0 from non-hit and text-mode traces. Gallery local-store-development aliases to 'local-store' so the master image satisfies pkg/model.LocalStoreBackend lookups from the embedding cache. Misc: llama-cpp TokenizeString reads the correct 'prompt' JSON key (the original bug); ModelTokenize nil-guard; non-fatal mitm proxy startup; PII 'route_local' renamed to 'allow' with docs/UI in sync; model-editor footer no longer eats the edit area on small screens; several config-editor template/dropdown/section fixes. Tests: e2e router specs (casual/code-hint + long-conversation trim), vector_store trace specs, lazy-factory specs, gallery dev-alias resolution, Playwright trace badge + scroll regression. Assisted-by: Claude:claude-opus-4-7 [Claude Code] Signed-off-by: Richard Palethorpe <io@richiejp.com> * feat(backend): auto-size batch to context for embedding and rerank models Embedding and rerank models pool over the whole input in a single physical batch (n_ubatch). With batch left at the 512 default, the backend rejects longer inputs with "input is too large to process", silently capping a large-context embedder (e.g. 8k/32k) at 512 tokens. Size n_batch to the context for these single-pass usecases, mirroring the existing FLAG_SCORE behaviour; an explicit batch: still wins. Extracts EffectiveContextSize/EffectiveBatchSize from grpcModelOpts so the effective decode window has one home for other callers to reuse. Adds an e2e-aio regression test that embeds a >512-token input. The AIO embedding model is switched to nomic-embed-text-v1.5 (2048 context) because the previous granite model was capped at 512 tokens and could not exercise the larger batch. Assisted-by: claude-code:claude-opus-4-8 [Claude Code] Signed-off-by: Richard Palethorpe <io@richiejp.com> * fix(gallery): raise arch-router scoring output cap via parallel:64 Scoring decodes the whole prompt+candidate in a single llama_decode and reads one logit row per candidate token. The vendored llama.cpp server caps causal output rows at n_parallel, so the default of 1 aborts with GGML_ASSERT(n_outputs_max <= cparams.n_outputs_max) on multi-token route labels. Set options: [parallel:64] on both arch-router quant entries to lift the cap; kv_unified (the grpc-server default) keeps the full context per sequence, so this does not split the KV cache. Assisted-by: claude-code:claude-opus-4-8 [Claude Code] Signed-off-by: Richard Palethorpe <io@richiejp.com> --------- Signed-off-by: Richard Palethorpe <io@richiejp.com>
46 lines
1.8 KiB
Go
46 lines
1.8 KiB
Go
package localaitools
|
|
|
|
import (
|
|
"context"
|
|
|
|
"github.com/modelcontextprotocol/go-sdk/mcp"
|
|
)
|
|
|
|
func registerPIITools(s *mcp.Server, client LocalAIClient, _ Options) {
|
|
mcp.AddTool(s, &mcp.Tool{
|
|
Name: ToolListPIIPatterns,
|
|
Description: "List the active PII regex pattern set. Each entry shows the pattern id, description, and current action (mask, block, allow). Read-only.",
|
|
}, func(ctx context.Context, _ *mcp.CallToolRequest, _ struct{}) (*mcp.CallToolResult, any, error) {
|
|
patterns, err := client.ListPIIPatterns(ctx)
|
|
if err != nil {
|
|
return errorResult(err), nil, nil
|
|
}
|
|
return jsonResult(patterns), nil, nil
|
|
})
|
|
|
|
mcp.AddTool(s, &mcp.Tool{
|
|
Name: ToolGetPIIEvents,
|
|
Description: "Recent PII redaction events. Filter by correlation_id (joins to a usage record), user_id, or pattern_id. Events never carry the matched value — only an 8-char sha256 prefix so admins can dedupe recurring leaks.",
|
|
}, func(ctx context.Context, _ *mcp.CallToolRequest, args PIIEventsQuery) (*mcp.CallToolResult, any, error) {
|
|
events, err := client.GetPIIEvents(ctx, args)
|
|
if err != nil {
|
|
return errorResult(err), nil, nil
|
|
}
|
|
return jsonResult(events), nil, nil
|
|
})
|
|
|
|
mcp.AddTool(s, &mcp.Tool{
|
|
Name: ToolTestPIIRedaction,
|
|
Description: "Dry-run the PII redactor against text without recording a real event. Useful for tuning patterns: paste a candidate string and see whether it would be masked, blocked, or routed locally.",
|
|
}, func(ctx context.Context, _ *mcp.CallToolRequest, args PIIRedactTestRequest) (*mcp.CallToolResult, any, error) {
|
|
if args.Text == "" {
|
|
return errorResultf("text is required"), nil, nil
|
|
}
|
|
res, err := client.TestPIIRedaction(ctx, args)
|
|
if err != nil {
|
|
return errorResult(err), nil, nil
|
|
}
|
|
return jsonResult(res), nil, nil
|
|
})
|
|
}
|