create-branch

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local file diffs to generate branch names, creating a surface for potential instructions embedded within the codebase to influence agent behavior.
  • Ingestion points: The skill reads repository state via git diff, git diff --cached, and git status --short in SKILL.md.
  • Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the analyzed diff content.
  • Capability inventory: The skill has the ability to execute shell commands including git checkout -b, git checkout, and gh api as defined in SKILL.md.
  • Sanitization: The skill instructs the agent to transform the description into a kebab-case, ASCII-only string and requires user confirmation before execution.
  • [COMMAND_EXECUTION]: The skill executes multiple shell commands to interact with the Git version control system and GitHub CLI.
  • Evidence: Uses gh api user --jq .login to fetch the username and various git commands (Step 5) to identify default branches and create new branches.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 07:22 AM
Security Audit — agent-trust-hub — create-branch