riftri-worktrees
Pass
Audited by Gen Agent Trust Hub on Nov 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
riftriandgitCLI tools to perform repository operations. It includes specific instructions to avoid forced operations, unauthorized installations, and modifications to system configurations like shell profiles. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from external Git repositories. 1. Ingestion points: Repository state and command output (e.g., via
git statusandriftri doctor) processed in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Execution of riftri and git commands. 4. Sanitization: Absent, though the use of structured JSON output is encouraged for error handling.
Audit Metadata