github

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides instructions for the agent to use git and gh CLI tools to automate repository tasks.
  • Evidence: SKILL.md and references/stacked-pr-workflow.md contain templates for git rebase --onto, git push --force-with-lease, and gh pr merge.
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by interpolating external metadata into command-line arguments (Category 8).
  • Ingestion points: Branch names, pull request titles, and PR numbers are ingested from the repository environment in SKILL.md and references/stacked-pr-workflow.md.
  • Boundary markers: The skill instructions do not define boundary markers or include directives to ignore embedded instructions in the ingested content.
  • Capability inventory: The agent is instructed to perform subprocess execution of git and gh tools using these external strings.
  • Sanitization: There is no requirement or logic provided to sanitize or escape user-controlled strings like <PR title> before they are executed in a shell context.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 05:37 AM