chroma
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates retrieval-augmented generation (RAG), which creates an attack surface for indirect prompt injection through stored documents.\n
- Ingestion points: External data is ingested into the agent's context through document addition in SKILL.md (e.g., collection.add(documents=...)) and retrieval in references/integration.md.\n
- Boundary markers: Code examples do not demonstrate the use of delimiters or warnings to ignore embedded instructions in retrieved content.\n
- Capability inventory: The skill demonstrates database file-write operations for persistence (PersistentClient in SKILL.md) and network operations via its server mode (HttpClient in SKILL.md).\n
- Sanitization: No sanitization, escaping, or validation of document content is provided in the integration examples.\n- [EXTERNAL_DOWNLOADS]: The skill references standard installation of well-known libraries from official registries.\n
- Evidence: References installation of chromadb, sentence-transformers, and integration with langchain and llama-index frameworks.
Audit Metadata