The Agent Skills Directory

[COMMAND_EXECUTION] (LOW): The skill is authorized to use the Bash tool. While the instructions focus on documentation and analysis, the Bash tool inherently allows for arbitrary command execution. This risk is managed by the skill's strict 'Phase 3: APPROVAL GATE' which requires human consent before executing documentation plans.
[PROMPT_INJECTION] (LOW): The skill is susceptible to Indirect Prompt Injection (Category 8) because it ingests and processes untrusted external code.
Ingestion points: Files read through the Read, Grep, and Glob tools during documentation discovery.
Boundary markers: Absent; there are no instructions to ignore embedded commands within the code being documented.
Capability inventory: Bash, Write, Edit, and Read tools are available to the agent.
Sanitization: Absent; the skill does not specify escaping or filtering content extracted from the codebase.

documentation