implement-slice
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from project files and Git history to determine its actions.
- Ingestion points: Reads feature specifications from
spec/features/<slug>/slices.json, behavior definitions fromspec.md, and state information from Git commit trailers (Slice,Slice-Pending). - Boundary markers: The instructions do not define clear delimiters or warnings to ignore instructions embedded within these specification files.
- Capability inventory: The agent has the ability to write to the file system, perform Git operations (commit, push, branch management), and execute project-level scripts.
- Sanitization: There is no mention of sanitizing or validating the content of the specifications or Git trailers before using them to guide the build process.
- [COMMAND_EXECUTION]: The skill is designed to run arbitrary project code as part of its verification process.
- Execution: Step 6 explicitly instructs the agent to run the project's 'typecheck, lint and tests'.
- Context: This is the primary purpose of the skill and occurs within the user's project environment. No evidence of unauthorized or malicious command execution was found.
Audit Metadata