Subconscious semantic knowledge requires both a vector store

Error message

Subconscious semantic knowledge requires both a vector store and an embedder.

What it means

getKnowledgeSemanticIndex lazily creates a KnowledgeSemanticIndexCoordinator that performs semantic search over knowledge. Because this requires embedding knowledge into a vector store, it asserts that both `this.vector` and `this.embedder` exist and throws this combined error when either is missing. Unlike the constructor checks, this one fires lazily at call time via drainKnowledgeSemanticIndex or semanticCandidates.

Source

Thrown at packages/memory/src/index.ts:534

        throw new Error('Subconscious semantic knowledge requires a vector store. Pass a `vector` option to Memory.');
      }
      if (!this.embedder) {
        throw new Error('Subconscious semantic knowledge requires an embedder. Pass an `embedder` option to Memory.');
      }
    }
  }

  private async getKnowledgeStore(): Promise<KnowledgeStorage> {
    const store = await this.storage.getStore('knowledge');
    if (!store) {
      throw new Error(`Knowledge storage domain is not available on ${this.storage.constructor.name}`);
    }
    return store;
  }

  public async getKnowledgeSemanticIndex(): Promise<KnowledgeSemanticIndexCoordinator> {
    if (!this.vector || !this.embedder) {
      throw new Error('Subconscious semantic knowledge requires both a vector store and an embedder.');
    }
    this._knowledgeSemanticIndex ??= this.getKnowledgeStore().then(
      knowledge =>
        new KnowledgeSemanticIndexCoordinator({
          knowledge,
          vector: this.vector!,
          embedder: this.embedder!,
          embedderOptions: this.embedderOptions,
        }),
    );
    return this._knowledgeSemanticIndex;
  }

  public async drainKnowledgeSemanticIndex(scope?: KnowledgeScope): Promise<number> {
    return (await this.getKnowledgeSemanticIndex()).drain(scope);
  }

  /**

View on GitHub (pinned to 75dd419e61)

Solutions

Pass both `vector` and `embedder` options to the Memory constructor
Avoid calling drainKnowledgeSemanticIndex/semanticCandidates when semantic indexing is not configured
Guard with checks (memory has vector && embedder) before invoking semantic index APIs

Example fix

// before
await memory.drainKnowledgeSemanticIndex(); // Memory built without vector/embedder
// after
new Memory({ storage, vector: new LibSQLVector({ connectionUrl: url }), embedder: new FastEmbed(), ... });
await memory.drainKnowledgeSemanticIndex();

Defensive patterns

Strategy: try-catch

Validate before calling

if (!memoryHasVectorAndEmbedder(memory)) {
  console.warn('Skipping semantic knowledge indexing: vector/embedder not configured');
} else {
  await memory.drainKnowledgeSemanticIndex();
}

Type guard

function canSemanticIndex(m: Memory): boolean {
  return 'vector' in m && 'embedder' in m && Boolean((m as any).vector) && Boolean((m as any).embedder);
}

Try / catch

try {
  await memory.drainKnowledgeSemanticIndex();
} catch (err) {
  if (err instanceof Error && err.message.includes('requires both a vector store and an embedder')) {
    // degrade gracefully: skip semantic indexing or lazily configure vector+embedder
  } else throw err;
}

Prevention

Only call drainKnowledgeSemanticIndex/semanticCandidates on fully configured Memory instances
Gate semantic indexing behind a config check (vector && embedder present)
Construct Memory via a factory that guarantees vector+embedder when knowledge features are on
Add integration tests that exercise semantic indexing with real vector+embedder config

When it happens

Trigger: Calling drainKnowledgeSemanticIndex or semanticCandidates (which call getKnowledgeSemanticIndex) on a Memory that has knowledge storage but was constructed without a `vector` option or an `embedder` option (or either).

Common situations: Constructor-time subconscious validation skipped (feature enabled via merged thread config or a code path bypassing constructor checks), so the failure surfaces later at index time; partially configured Memory where only vector or only embedder was set.

Related errors

AI-assisted analysis of mastra-ai/mastra@75dd419e61 (2026-08-30). Data as JSON: /api/errors/695627848630f689. Report an issue: GitHub.