The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a Python script scripts/check-output.py designed to be executed by the agent to validate that generated deliverables contain required sections. The script uses the standard library to read a local file path and check for specific header strings. This is a common utility function for ensuring output quality and does not perform dangerous operations like network access or privilege escalation.
[DATA_EXPOSURE]: The script scripts/check-output.py reads a file from the local file system. While it accesses file content, it only checks for the presence of specific marketing-related headers and does not log, transmit, or otherwise expose the file's content.
[INDIRECT_PROMPT_INJECTION]: The skill has a small attack surface where scripts/check-output.py reads untrusted data (a user-generated deliverable). However, the script performs simple string matching and does not interpret or execute the content of the file, mitigating risk.

paid-advertising