The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Bash tool to perform extensive file system operations, searching for memory indices and calculating directory statistics within ~/.claude/ and .claude/context/.
[COMMAND_EXECUTION]: Invokes local Node.js functionality via node -e to call methods from the memory-manager.cjs module for API verification.
[PROMPT_INJECTION]: The skill processes untrusted content from memory files which creates a surface for indirect prompt injection. 1. Ingestion points: Reads MEMORY.md, learnings.md, and JSON data stores. 2. Boundary markers: Lacks explicit separators or instructions to ignore instructions embedded within the memory data. 3. Capability inventory: Uses Bash and TaskUpdate to perform actions and write to the file system based on audit results. 4. Sanitization: Does not perform validation or filtering of the content retrieved from memory stores before processing.

memory-audit