The Agent Skills Directory

[COMMAND_EXECUTION]: The skill incorporates a local Python script, scripts/check-output.py, which is intended for execution by the agent to validate the quality of its output. Analysis of the script confirms it only performs basic string matching and keyword verification against a local file path, with no capabilities for network access or system modification.
[INDIRECT_PROMPT_INJECTION]: The skill includes a validation process where a script reads the agent's generated deliverable. 1. Ingestion points: scripts/check-output.py reads a file containing agent-produced text based on user input. 2. Boundary markers: None present. 3. Capability inventory: The script uses Path.read_text() but contains no dangerous execution sinks such as eval or subprocess. 4. Sanitization: The script only performs case-insensitive keyword searches to verify term presence.

partner-programs