The Agent Skills Directory

[SAFE]: The skill provides documentation and examples for using ChromaDB, a standard vector database. No malicious behavior or unsafe code was detected.
[PROMPT_INJECTION]: The skill defines a surface for indirect prompt injection as it facilitates the ingestion of external documents into a vector database for retrieval. This is a functional requirement of the technology.
Ingestion points: collection.add, Chroma.from_documents, and VectorStoreIndex.from_documents (SKILL.md, references/integration.md).
Boundary markers: Not specified in the provided code examples.
Capability inventory: The skill documentation does not define specific executable tools, but the database stores content for agent retrieval.
Sanitization: No explicit sanitization or filtering logic is provided in the examples.
[EXTERNAL_DOWNLOADS]: The documentation references standard package installation from PyPI and NPM (e.g., chromadb, sentence-transformers). These references point to official repositories of well-known services.
[CREDENTIALS_UNSAFE]: API key references in the code snippets use generic placeholders (e.g., "your-key"), which is a safe and standard practice for documentation.

chroma