The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Python scripts (vector_search.py and memory_store.py) to manage its knowledge base. These scripts are invoked via standard command-line interfaces to perform searches and updates.
[DATA_EXFILTRATION]: The skill accesses a project memory database file at ~/mnt/.claude/.cowork-mem/memory.db. This path is used exclusively for internal state management and retrieval, with no evidence of unauthorized data transfer to external sources.
[PROMPT_INJECTION]: As a RAG (Retrieval-Augmented Generation) tool, the skill is susceptible to indirect prompt injection from data stored in the memory database.
Ingestion points: Project memory is read from memory.db via the search scripts.
Boundary markers: None explicitly defined in the retrieval instructions.
Capability inventory: The skill can read and write to the local database file using the provided Python scripts.
Sanitization: No specific sanitization or filtering of stored insights is described in the prompt logic.

knowledge-agent