The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill creates an ingestion surface for untrusted data that could influence the agent during the retrieval phase.
Ingestion points: The ingest tool in mcp/servers/rag-server/server.py accepts document content and metadata which may originate from external, untrusted sources.
Boundary markers: There are no boundary markers or instructions present to ensure the agent ignores potentially malicious instructions embedded within the retrieved data.
Capability inventory: The skill possesses network capabilities to communicate with a Qdrant database and handles data processing via local embedding models.
Sanitization: The implementation does not include sanitization or filtering logic to identify or remove prompt injection patterns from ingested text.

rag-builder