finishing-a-development-branch
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes standard development and version control commands including
npm test,cargo test,pytest,go test, and variousgitsubcommands (merge,branch,worktree,push). It also utilizes thegh(GitHub) CLI for Pull Request creation. These operations are consistent with the skill's stated purpose of finalizing implementation work. - [INDIRECT_PROMPT_INJECTION]: The skill identifies an attack surface where untrusted data from the local environment is ingested into the agent context.
- Ingestion points: The agent is instructed to read and display test suite failures (Step 1) and Git branch lists (Step 2, Step 5).
- Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore instructions that might be embedded in test output or commit logs.
- Capability inventory: The skill possesses capabilities for shell command execution, file system cleanup (via
git worktree remove), and remote interaction (viagit pushandgh pr create). - Sanitization: The skill uses a quoted heredoc (
cat <<'EOF') when generating the Pull Request body in Step 4, which is a best practice to prevent unexpected shell interpolation of the generated content.
Audit Metadata