The Agent Skills Directory

[SAFE]: The skill provides a structured methodology for project risk analysis without any dangerous capabilities.
[COMMAND_EXECUTION]: The provided Python script scripts/risk_categorizer.py is a local utility tool that processes user-supplied JSON data. Analysis of the source code confirms it only uses standard Python libraries (argparse, json, sys) and performs no network requests, file system writes, or subprocess executions.
[INDIRECT_PROMPT_INJECTION]: While the skill processes external data through the risk_categorizer.py script, it lacks exploitable capabilities such as network access or the ability to write to the file system. The script performs validation on input fields and provides controlled output to the console.
[METADATA_POISONING]: YAML frontmatter and skill metadata fields are consistent with the documented project management purpose and do not contain deceptive or malicious instructions.

pre-mortem