write-internal-docs
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the Bash tool to perform essential repository tasks, including cloning the documentation repository, managing branches, and pushing updates. It also utilizes the GitHub CLI (
gh) to create pull requests. All operations are restricted to the vendor's official repository (browseros-ai/internal-docs). - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by reading source code and existing documentation files to generate its content. This activity is a core part of its documentation function but inherently introduces a vulnerability surface where malicious content in those files could attempt to influence the agent's output.
- Ingestion points: The agent reads local project files and repository templates (e.g., in
.internal-docs/) during the research and drafting phases. - Boundary markers: The instructions do not define specific delimiters or instructions to ignore instructions embedded within the files being researched.
- Capability inventory: The agent has access to the Bash tool for file system and repository operations, along with general Read and Write permissions.
- Sanitization: The workflow includes a step to remove YAML frontmatter when rendering HTML, but does not provide specific sanitization for the narrative content of the documentation.
Audit Metadata