docs
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
devtools::document()when working with R projects. While this is a standard development tool, it involves executing environment-specific commands to generate documentation files. - [INDIRECT_PROMPT_INJECTION]: The skill has a vulnerability surface for indirect prompt injection because it processes untrusted content from the codebase to analyze and fix documentation.
- Ingestion points: The agent reads source code files and documentation (e.g.,
CLAUDE.md) within the specified scope (staged files, PR changes, or the entire codebase) as defined in Phases 1 and 2. - Boundary markers: No explicit delimiters or instructions to ignore embedded commands within the analyzed code are provided.
- Capability inventory: The skill grants the agent file-write capabilities to generate and update documentation files across the repository in Phase 3.
- Sanitization: There is no evidence of sanitizing or escaping the codebase content before it is processed by the agent's logic.
- [PROMPT_INJECTION]: The skill contains a control instruction ('automatically exit after completing all phases successfully') triggered by the 'IMPORTANT' marker. This is intended to modify the agent's default operational flow and response behavior.
Audit Metadata