The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data that could contain malicious instructions designed to influence the agent in future sessions. 1. Ingestion points: Reads existing memory files from the project directory and accepts user-provided insight strings. 2. Boundary markers: The skill does not implement delimiters or instructions to ignore embedded commands within the ingested content. 3. Capability inventory: The skill has read and write access to the project's memory file system. 4. Sanitization: No escaping, validation, or sanitization of user-provided insights is performed before they are integrated into persistent memory files.

remember-insight