ce-help
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data that could theoretically contain instructions. * Ingestion points: user messages, project configuration files (.compound-engineering/config.yaml), and external documentation fetched via HTTP. * Boundary markers: No specific delimiters or safety instructions are defined to isolate external data. * Capability inventory: The skill is authorized to read files, execute read-only git commands, and perform network requests. * Sanitization: No explicit sanitization or validation of the ingested content is described.
- [EXTERNAL_DOWNLOADS]: The skill fetches documentation and guidelines from the author's official GitHub repository (EveryInc/compound-engineering-plugin). These are legitimate vendor resources used for providing help.
- [COMMAND_EXECUTION]: Instructions direct the agent to run read-only git and pull request checks to determine repository status and workflow progress.
Audit Metadata