ck-plan
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from the user's repository to generate implementation plans, which creates a potential surface for indirect prompt injection.
- Ingestion points: The skill reads repository content, symbol definitions, and command results in
SKILL.mdto ground its implementation plans. - Boundary markers: While the skill establishes a strict "Plan and coverage contract" for its workers to follow, it does not specify explicit delimiters or sanitization for the content ingested from the repository.
- Capability inventory: The skill utilizes
invoke-harness, fresh native sub-agents, repository exploration tools, and the execution of repository-based checks and tests. - Sanitization: The instructions do not define specific sanitization or filtering logic for repository-sourced data.
- [COMMAND_EXECUTION]: The planning and review cycle explicitly involves executing "repository-supported fast checks" and reporting "command results" to verify the feasibility and correctness of the proposed implementation plan.
Audit Metadata