The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and distill information from project files, conversation history, and external memory stores (such as ctx_memory and claude-mem). This ingestion process creates a potential surface for indirect prompt injection, where malicious instructions embedded in project data could be persisted into the documentation, potentially affecting future agent sessions. The skill mitigates this by instructing the agent to distinguish between active norms, observed behavior, and unverified claims.
[COMMAND_EXECUTION]: The instructions require the agent to run the strictdoc export . command to validate documentation and configuration changes. This involves the execution of an external tool within the local environment.
[DYNAMIC_EXECUTION]: The skill includes a configuration file strictdoc_config.py. As is standard for the StrictDoc tool, this Python script is intended to be loaded and executed to define project settings, representing the execution of provided script content.

save-mem