write-documentation
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute code examples and run repository-specific commands (such as format and doc-lint) resolved from the project's AGENTS.md file.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it reads project files and executes code examples found within them. 1. Ingestion points: Project source code, existing documentation, and task requirements are read into the agent context (SKILL.md). 2. Boundary markers: The instructions do not specify any markers or warnings to ignore malicious content within the processed files. 3. Capability inventory: The agent is authorized to run arbitrary code examples and maintenance tools (SKILL.md Rules 3 and 8). 4. Sanitization: There are no instructions to sanitize or validate the code examples or command strings before execution.
Audit Metadata