Skip to content
Agentmart
Submit a tool

ticket-ship

SHIP stage of /start - writes one report with numbered questions, takes the user's answers, runs round 2 if they ask for changes, and opens a draft PR only on…

Skill by beefysalad

IndexedMIT
Write a review

Chalk Labs

Need help deploying ticket-ship?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/beefysalad-ticket-ship

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

SHIP stage of /start - writes one report with numbered questions, takes the user's answers, runs round 2 if they ask for changes, and opens a draft PR only on their explicit go-ahead after a secrets scan.

Use cases: documents, security

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md and code. Checked Nov 2, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/ticket-ship

What's inside

Skill name
ticket-ship
Description
SHIP stage of /start - writes one report with numbered questions, takes the user's answers, runs round 2 if they ask for changes, and opens a draft PR only on their explicit go-ahead after a secrets scan.
License
MIT
Instructions
32 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# SHIP

Read `patrick-workflows:ticket-workspace` first. `S` = `bash "$SKILL_DIR/../ticket-workspace/scripts/state.sh" <WS>/state.md`.

## 0. Every entry
`git branch --show-current` must equal `branch`; if not and the tree is clean, `git switch <branch>`; if dirty, stop and tell the user. Never push any other branch.

## 1. Write the report (`WS/handoff.md`), in this order
1. **Outcome:** `READY` or `BLOCKED: <reason>`; depth; tasks done / total; dispatches used / budgets. The PR title type is the `type` state key (absent in workspaces from before v0.4.0: `fix` when the branch starts with `fix/`, else `feat`).
2. **Needs your decision:** numbered questions, each with your recommendation, answerable in a few words ("1 keep 2 change: ... 3 yes"). Include deferred Importants, out-of-scope files, unreproduced findings that need a call, blockers, and base drift. The last question is always "Open the draft PR? (yes / no)".
3. **Unreproduced findings** (security first). Then **Refuted by the challenger** (depth full): copy that section of `gauntlet-review/report.md` (and `gauntlet-review-r2/report.md`), Criticals first; each refuted Critical is also a question under "Needs your decision" ("Reopen F1-n? recommended: <keep closed | reopen>"). A reopened finding becomes an Important or Critical for round 2.
4. **Severity downgrades.**
5. **What changed:** `git diff --stat <base>..HEAD`, `git log --oneline <base>..HEAD`, scope classes from the last scope check.
6. **Evidence:** gates vs baseline (known reds called out); red→green per task from `reports/task-N.md`; review-loop verdicts per round, graded scores per round and the exit-pair result from `gauntlet-review/` (and `gauntlet-review-r2/` after a round 2).
7. **Rulings:** every entry of `rulings.md`, with cost if wrong.
8. **Base drift:** if a remote exists, `git fetch origin <base_branch> -q` then `git rev-list --count <base>..origin/<base_branch>`. Non-zero → "Base moved by N commits; rebase before the PR?" is one of the questions.
9. **Draft PR:** title `<type>(<id>): <title>`; body in `WS/pr-body.md` with Summary, Changes, How it was tested (the evidence), and the ticket reference. No tool attribution of any kind.

`S phase handoff`. Show sections 1–2 in chat with the path to `handoff.md`, then wait for the answers.

## 2. Answers
- Record each answer as a ruling with `source: user`.
- An answer that asks for a code change: append the needed tasks to `plan.md` under `## Round 2` (in the plan's task format, with tests, numbered from `tasks_total` + 1 so they are never mistaken for finished round-1 tasks), `S set round2 yes`, add their count to `tasks_total`, `S set budget_impl_max` to `budget_impl_used` + 4 × new tasks (2 × for lite), `S set preflight no`, `S phase round2`, and invoke `patrick-workflows:ticket-build`. Its review loop gets a fresh `--budget` and runs its bars on the whole branch.
- "Rebase": `git rebase origin/<base_branch>`. On conflict: `git rebase --abort` and ask the user how to proceed. After a clean rebase, `S set base $(git merge-base HEAD origin/<base_branch>)` so every later diff covers only this branch, rerun the gates once, note in the report that the baseline predates the rebase, and say so in the PR body.

## 3. Open the PR (only after an explicit yes)
1. Secrets: `git diff <base>..HEAD | bash "$SKILL_DIR/../ticket-workspace/scripts/secret-scan.sh" [--deny-file .claude/patrick-workflows-deny.txt] - "<WS>/pr-body.md"`; pass `--deny-file` only when the project has that file. Any hit → show the rule and location (never the value) and stop: this is security-sensitive and the user decides.
2. Show the final PR title and body.
3. `S set push_approved yes` (the guard hook blocks a ticket's push without it), then `git push -u origin <branch>`, then `gh pr create --draft --base <pr_target> --title "<title>" --body-file "<WS>/pr-body.md"`.
4. `S set pr_url <url>`, `S phase pr`. Give the user the URL and say that after the PR is merged they should run `/start <id>` once more to mark the ticket done and clean up.

From beefysalad/patrick-workflows, MIT.

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Oct 5, 2026 (1 month ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from beefysalad

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get ticket-ship running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying ticket-ship?

CallGet help
Cite this page

ticket-ship skill. Agent Mart. https://agentmart.dev/skills/beefysalad-ticket-ship (last updated 2026-10-24)