The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill executes inline Python scripts using python3 -c to process and format JSON output from the audit tool. These scripts are statically defined within the skill instructions.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and acts upon data generated by an external tool (fallow audit), which could theoretically include maliciously crafted code symbols if the scanned codebase is compromised.
Ingestion points: Output from the npx fallow audit command in SKILL.md.
Boundary markers: None specified to delimit tool output from agent instructions.
Capability inventory: The skill can modify files (refactoring), delete files (delete-file), and execute shell commands (npx, python3).
Sanitization: The tool output is structured as JSON and filtered through a Python script before being presented to the agent, providing a layer of structural validation.
[EXTERNAL_DOWNLOADS]: The skill utilizes npx to run the fallow package. This is a standard mechanism for executing Node.js-based tools provided by the vendor.
[COMMAND_EXECUTION]: The skill runs shell commands to perform audits and process data, including npx fallow and python3 invocations.

improve-diff