The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of code files and git diffs to identify comments and code structures for deletion.
Ingestion points: Untrusted code and diffs are loaded from the caller's files or the repository state (SKILL.md, Scope section).
Boundary markers: The instructions lack explicit boundary markers or instructions to ignore embedded prompts within the code being analyzed.
Capability inventory: The skill can delete code paths, modify application logic, and trigger the execution of other skills such as /how, /why, and /architect based on the analysis of input code.
Sanitization: There is no evidence of sanitization or escaping of the code content before it is used to influence agent decisions and tool invocations.
[COMMAND_EXECUTION]: The skill performs direct modifications to the filesystem and invokes additional internal tools.
It performs destructive operations such as deleting code paths and removing parameters (SKILL.md, Step 3).
It invokes multiple sub-skills and spawns worker agents to process the code (SKILL.md, Steps 1 and 2).

no-comments