fix(cli): resolve embedding provider explicitly and surface lane status in sl search (#192)

* feat(cli): add tryUseManagedLocalEmbeddingsDaemon for read-only callers

* feat(cli): add resolveProjectEmbeddingProvider helper

* fix(cli): wire sl search through resolveProjectEmbeddingProvider so semantic lane works

* fix(cli): wire wiki/knowledge search through resolveProjectEmbeddingProvider

* feat(cli): surface embeddings-unavailable status when sl search returns empty

* refactor(cli): route admin reindex through resolveProjectEmbeddingProvider

* refactor: pass embeddingProvider into ingest/scan instead of resolving inside @ktx/context

* refactor(mcp): resolve embedding provider in CLI factory, pass into context ports

* refactor(context): delete MANAGED_SENTENCE_TRANSFORMERS_BASE_URL sentinel

* refactor(cli): delete sentinel-based managed-embeddings indirection

* chore: scrub stale managed-embeddings sentinel references from tests and smoke script

* chore: unexport unused EmbeddingResolutionMode alias

* fix(cli): force pathPrefix="" when targeting the managed embeddings daemon

The managed daemon serves /embeddings/compute directly. The default
pathPrefix in @ktx/llm is /api, so omitting sentenceTransformers from
ktx.yaml produced /api/embeddings/compute -> 404. The resolver now
sets pathPrefix='' explicitly when wiring the managed daemon URL,
matching what the daemon actually exposes.
This commit is contained in:
Andrey Avtomonov 2026-05-21 02:21:22 +02:00 committed by GitHub
parent 56a967278a
commit 9d92c79988
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
36 changed files with 750 additions and 442 deletions

View file

@ -8,7 +8,6 @@ import { noopLogger, SessionWorktreeService } from '../core/index.js';
import type { KtxSemanticLayerComputePort } from '../daemon/index.js';
import {
createRuntimeToolDescriptorFromAiTool,
createLocalKtxEmbeddingProviderFromConfig,
createLocalKtxLlmRuntimeFromConfig,
KtxIngestEmbeddingPortAdapter,
RuntimeAgentRunner,
@ -16,6 +15,7 @@ import {
type KtxLlmRuntimePort,
type KtxRuntimeToolSet,
} from '../llm/index.js';
import type { KtxEmbeddingProvider } from '@ktx/llm';
import type { KtxLocalProject } from '../project/index.js';
import { ktxLocalStateDbPath } from '../project/index.js';
import { PromptService } from '../prompts/index.js';
@ -114,6 +114,7 @@ export interface CreateLocalBundleIngestRuntimeOptions {
queryExecutor?: KtxSqlQueryExecutorPort;
jobIdFactory?: () => string;
logger?: KtxLogger;
embeddingProvider?: KtxEmbeddingProvider | null;
}
export interface LocalBundleIngestRuntime {
@ -669,7 +670,7 @@ export function createLocalBundleIngestRuntime(
mkdirSync(join(options.project.projectDir, '.ktx/cache/local-ingest'), { recursive: true });
const store = new SqliteBundleIngestStore({ dbPath });
const contextStore = new SqliteContextEvidenceStore({ dbPath });
const embeddingProvider = createLocalKtxEmbeddingProviderFromConfig(options.project.config.ingest.embeddings);
const embeddingProvider = options.embeddingProvider ?? null;
const embedding = embeddingProvider ? new KtxIngestEmbeddingPortAdapter(embeddingProvider) : new NoopEmbeddingPort();
const connections = new LocalConnectionCatalog(options.project, options.queryExecutor);
const rootFileStore = options.project.fileStore;

View file

@ -34,6 +34,7 @@ export interface RunLocalIngestOptions {
semanticLayerCompute?: KtxSemanticLayerComputePort;
queryExecutor?: KtxSqlQueryExecutorPort;
logger?: KtxLogger;
embeddingProvider?: import('@ktx/llm').KtxEmbeddingProvider | null;
}
export interface LocalIngestMcpOptions
@ -172,6 +173,7 @@ async function runScheduledPullJob(options: {
semanticLayerCompute?: KtxSemanticLayerComputePort;
queryExecutor?: KtxSqlQueryExecutorPort;
logger?: KtxLogger;
embeddingProvider?: import('@ktx/llm').KtxEmbeddingProvider | null;
}): Promise<LocalIngestResult> {
const runtime = createLocalBundleIngestRuntime(options);
const jobId = options.jobId ?? runtime.nextJobId();
@ -225,6 +227,7 @@ export async function runLocalIngest(options: RunLocalIngestOptions): Promise<Lo
semanticLayerCompute: options.semanticLayerCompute,
queryExecutor: options.queryExecutor,
logger: options.logger,
embeddingProvider: options.embeddingProvider,
});
}
@ -403,6 +406,7 @@ export async function runLocalMetabaseIngest(
semanticLayerCompute: options.semanticLayerCompute,
queryExecutor: options.queryExecutor,
logger: options.logger,
embeddingProvider: options.embeddingProvider,
});
} catch (error) {
child = await recordLocalMetabaseChildFailure({