skills/mritd/skills/commit/Gen Agent Trust Hub

commit

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Performs git operations including status checks, staging, diff reading, and committing.
  • [COMMAND_EXECUTION]: References a vendor-specific tool, gitflow-toolkit, and provides a command for its installation if the necessary git ci alias is absent.
  • [PROMPT_INJECTION]: The skill uses git diff to analyze code changes, creating a surface for indirect prompt injection. 1. Ingestion points: Code changes read via git diff --cached. 2. Boundary markers: Absent. 3. Capability inventory: Command execution and file system access. 4. Sanitization: The diff content is processed without filtering.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 03:14 PM