merge-dependabot-prs

Pass

Audited by Gen Agent Trust Hub on Nov 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill makes extensive use of the GitHub CLI (gh) and git via the Bash tool to interact with repositories. This includes fetching metadata, merging pull requests, and approving workflow runs. While these are high-privilege operations, the skill instructions include mandatory user confirmation steps for risky actions such as bypassing branch protection or merging major version updates.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from GitHub pull requests, including titles, bodies, and commit messages. Specifically, in Step 5 ('Assess the majors'), the agent is instructed to read release notes and changelogs to determine the risk of an upgrade. This creates a surface for indirect prompt injection where an attacker could embed malicious instructions in a pull request description or release notes.
  • Ingestion points: The script scripts/dependabot_prs.py (in print_notes and details_blocks) extracts text from the PR body and commit messages. SKILL.md instructs the agent to read these extracted notes during assessment.
  • Boundary markers: Absent. The instructions do not define clear delimiters or specific warnings to ignore instructions embedded within the processed PR content.
  • Capability inventory: The agent has the ability to merge PRs (gh pr merge), comment on PRs (gh pr comment), and approve workflow runs (gh api ... approve).
  • Sanitization: The dependabot_prs.py script performs basic HTML-to-text conversion and strips detail blocks, but it does not perform semantic validation of the content, leaving it to the LLM to interpret safely.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 9, 2026, 03:12 PM