The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various local commands to manage the development lifecycle.
Runs project test suites using npm test, cargo test, pytest, or go test.
Executes multiple git commands for merging, pushing, and cleaning up worktrees.
Uses gh (GitHub CLI) to create Pull Requests and query user metadata.
[EXTERNAL_DOWNLOADS]: The 'Rescue: leaked secret' section suggests installing an external tool.
Recommends installing git-filter-repo via brew install or pip install to facilitate history rewriting.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface in Step 4 (Option 2).
Ingestion points: Reads local file content using cat to populate a Pull Request body.
Boundary markers: Uses EOF markers for the heredoc but does not include explicit instructions to the agent to ignore any embedded malicious content within the read files.
Capability inventory: Includes file reading (cat), network operations (git push, gh pr create), and command execution (git).
Sanitization: No specific sanitization or validation of the file content is performed before it is interpolated into the PR creation command.
[DYNAMIC_EXECUTION]: The history rewriting recipe uses the --email-callback flag of git-filter-repo.
This involves passing a Python expression as a string which is executed by the tool at runtime to process commit metadata.

finishing-a-development-branch