ce-plan
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a robust planning workflow with built-in security considerations for handling untrusted data and ensuring user oversight at critical decision points.
- [COMMAND_EXECUTION]: Executes repository-level git commands to determine project root, history, and metadata. This logic is restricted to local filesystem inspection for context-gathering purposes.
- [EXTERNAL_DOWNLOADS]: Employs dedicated agents to retrieve external documentation and Slack discussions. The skill includes specific defensive logic to treat all fetched content as data and ignore any instructions or code patterns found therein.
- [PROMPT_INJECTION]: The design proactively mitigates indirect prompt injection risks by explicitly instructing research agents to ignore agent-related directives contained within processed web or Slack content.
- [DATA_EXFILTRATION]: Includes a user-authorized feature to publish generated plans to an external Proof editor. This functionality is transparent, menu-driven, and intended for sharing finalized documentation.
Audit Metadata