spec-kitty-review
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Python scripts (".kittify/scripts/tasks/tasks_cli.py") and uses the "spec-kitty" CLI tool to automate task state transitions and code review workflows. These actions are performed within the local environment to maintain project status.
- [PROMPT_INJECTION]: The skill processes external data such as code diffs and work packages, which introduces an indirect prompt injection surface. 1. Ingestion points: Code diffs (git diff) and work package prompts mentioned in "SKILL.md" and "AUGMENTED.md". 2. Boundary markers: Absent in the provided skill files. 3. Capability inventory: The skill can execute local task-management scripts and CLI commands. 4. Sanitization: Absent from the documented workflow. The risk is minimized by the skill's specific focus on task-lane transitions.
Audit Metadata