The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill implements retrieval-augmented generation (RAG) patterns where external data is retrieved and injected into the agent's context window.
Ingestion points: Data is ingested from markdown files in the memory/ directory, rows in a SQLite database (agent-memory.db), and documents stored in a ChromaDB vector collection.
Boundary markers: The instructions recommend showing the source of each memory to help the agent judge its context, though the provided code snippets do not implement explicit delimiters or 'ignore' instructions for the retrieved data.
Capability inventory: The retrieved text is intended to be appended to system prompts or instruction files, which allows external content to influence the agent's behavior and responses.
Sanitization: The code templates do not include sanitization or validation of the memory content before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill references standard, well-known libraries for its functionality, including better-sqlite3, openai, and chromadb (PyPI).

agent-memory