ship_skill
- Shell
31
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
3 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill steveclarke/dotfiles --skill ship- SKILL.md6.6 KB
Overview
This skill runs the full pre-merge pipeline and creates a pull request in one command. It sequences simplify, code review, adversarial review (optional), finalize, documentation updates, shows a readiness dashboard, and then opens the PR if the branch is cleared. Use flags to show status only or force adversarial review.
How this skill works
The skill detects project context (project name, current branch, base branch) and skips if you're on the base. It runs a readiness check before each pipeline step; clean steps are skipped, failing steps invoke their respective skills and are logged. After all steps it prints a readiness dashboard and either creates a PR or reports blocking issues for user decision.
When to use it
- When you want a single command to run the full pre-merge checklist and create a PR
- When preparing a feature branch for review and merge
- To get an automated readiness dashboard before opening a PR
- When you want to enforce lint, tests, reviews, and docs updates consistently
- When you need to run an adversarial review for large or risky changes
Best practices
- Run from a feature branch, not the base branch; the skill aborts on base
- Push frequently so code review and finalize steps have current diffs
- Use --status to view the readiness dashboard without executing steps
- Pass --adversarial for sensitive or large changesets to force extra review
- Fix reported failures and re-run; completed steps are skipped automatically
Example use cases
- I finished a feature and want to run simplify, tests, docs checks, then open a PR
- Quickly check project readiness with --status before manual QA
- Force adversarial review on a large refactor with --adversarial
- Recover after cancelling a previous run; completed steps are cached and skipped
- Create a concise PR body that summarizes readiness results and test plan
FAQ
Shows the readiness dashboard only and stops; it does not run pipeline steps.
When does adversarial review run?
It runs if you pass --adversarial or if the diff exceeds the configured large-change threshold (default ~500 lines).
What happens if tests or code review fail?
The failing step is logged as failed, the pipeline stops, the dashboard is shown, and you must fix issues before a normal ship. You can choose to force PR creation if you explicitly confirm.