documentation
Fail
Audited by Gen Agent Trust Hub on May 20, 2026
Risk Level: CRITICALEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses shell commands like git, grep, and find to detect documentation drift and manage repository configuration, which is consistent with its auditing purpose.
- [EXTERNAL_DOWNLOADS]: It references reputable documentation frameworks like Nextra and Docusaurus and employs npx to execute linting tools like markdownlint and lychee. References to nextra.site, though flagged by external scanners, represent well-known services.
- [PROMPT_INJECTION]: The skill reads documentation and code files from the repository to identify drift, creating a surface for indirect prompt injection. Ingestion points include local files like CLAUDE.md and README.md, while capabilities involve repository modifications based on this data. No explicit sanitization or boundary markers are documented, which is standard for developer-facing tools.
Recommendations
- CRITICAL: 2 infected file(s) detected - DO NOT USE
- Contains 3 malicious URL(s) - DO NOT USE
Audit Metadata