The Agent Skills Directory

[SAFE]: The skill processes user-provided plan files but incorporates multiple safety checkpoints to mitigate risks associated with untrusted data ingestion.
Ingestion points: External implementation plan files.
Boundary markers: No specific technical delimiters are used.
Capability inventory: Involves executing tasks, running verifications, and using tools like update_plan based on plan content.
Sanitization: Relies on a mandatory critical review step by the agent and human-in-the-loop feedback loops after each batch.

executing-plans