issue-driven-github-flow

Pass

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses git and gh (GitHub CLI) to perform repository operations. The scripts/gitflow.sh script facilitates these tasks with validation for branch naming and commit messages.
  • [PROMPT_INJECTION]: The skill handles external data from GitHub, creating a surface for indirect prompt injection. 1. Ingestion points: Issue bodies and comments are retrieved via gh issue view. 2. Boundary markers: The workflow relies on task-specific sub-agent prompts. 3. Capability inventory: The skill can execute shell commands and modify code. 4. Sanitization: Inputs like branch names are checked via regex. The multi-agent review process (Planner/Reviewer/Implementer) provides a significant layer of defense-in-depth.
  • [SAFE]: Implements a robust 'no-code-without-plan' discipline, where implementation only begins after a formal approval verdict is recorded.
  • [SAFE]: Includes comprehensive security guidance in references/security.md regarding secret hygiene and push protection.
  • [SAFE]: Uses quoted heredocs in shell scripts to ensure data integrity and prevent injection when passing strings to CLI tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 17, 2026, 02:03 PM
Security Audit — agent-trust-hub — issue-driven-github-flow