ship
Skill3 starsChanged 23 days agoMITView on GitHub
Inspect, split into logical commits, validate, then deliver via scripts/ship.ps1 (direct push to main, or branch + PR when main requires it).
- Reads credentials
- Commits and pushes
194 other projects use a version of this.
Adopted by
Most recently taken up by
- Robo-Co-op/robobuilder-standardtheir own version, first seen Nov 16, 2026
- zyggy-org/zyggytheir own version, first seen Nov 16, 2026
- champ18ion/claud18codetheir own version, first seen Nov 16, 2026
- joelhooks/uncomplecttheir own version, first seen Nov 16, 2026
- claudialnathan/skillstheir own version, first seen Nov 15, 2026
- saleem-hadad/skillstheir own version, first seen Nov 15, 2026
Used with
Projects with it also use ios-sync, open-gstack-browser and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
Saqlain-811Skillios-sync
Regenerate the iOS debug bridge against the latest upstream gstack templates. (gstack)
Also in 6 projects that use this
seek888Skillopen-gstack-browser
Launch GStack Browser — AI-controlled Chromium with the sidebar extension baked in. Opens a visible browser window where you can watch every action in real time. The sidebar shows a live activity feed and chat. Anti-bot stealth built in. Use when asked to "open gstack browser", "launch browser", "connect chrome", "open chrome", "real browser", "launch chrome", "side panel", or "control my browser". Voice triggers (speech-to-text aliases): "show me the browser".
Also in 10 projects that use this
ayudaySkillsetup-deploy
Configure deployment settings for /land-and-deploy. Detects your deploy platform (Fly.io, Render, Vercel, Netlify, Heroku, GitHub Actions, custom), production URL, health check endpoints, and deploy status commands. Writes the configuration to CLAUDE.md so all future deploys are automatic. Use when: "setup deploy", "configure deployment", "set up land-and-deploy", "how do I deploy with gstack", "add deploy config".
Also in 19 projects that use this
ShreyRaviSkilllanding-report
Read-only queue dashboard for workspace-aware ship. Shows which VERSION slots are currently claimed by open PRs, which sibling Conductor workspaces have WIP work likely to ship soon, and what slot /ship would pick next. No mutations — just a snapshot. Use when asked to "landing report", "what's in the queue", "show me open PRs", or "which version do I claim next". (gstack)
Also in 9 projects that use this
zebbernSkilldebug-fix
Find and fix a bug or issue — from any source (GitHub issue, error message, user report, or observed behavior)
Also in 4 projects that use this
ramybarsoumSkilldocument-release
Post-ship documentation update. Reads all project docs, cross-references the diff, updates README/ARCHITECTURE/CONTRIBUTING/CLAUDE.md to match what shipped, polishes CHANGELOG voice, cleans up TODOS, and optionally bumps VERSION. Use when asked to "update the docs", "sync documentation", or "post-ship docs". Proactively suggest after a PR is merged or code is shipped.
Also in 23 projects that use this
Other versions194 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 194
- garrytan/gstack★ 142ksame name
- rtk-ai/rtk★ 87ksame name
- simstudioai/sim★ 30ksame name
- millionco/react-doctor★ 15ksame name
- nightly-labs/openbot★ 7.9ksame name
- BuilderIO/agent-native★ 7.7ksame name
- fallow-rs/fallow★ 5.4ksame name
- puncsky/system-design-and-architecture★ 3.6ksame name
- plugin87/ux-ui-agent-skills★ 3.2ksame name
- ShenSeanChen/waku-agent★ 2.2ksame name
184 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository12 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .cursor/rules/00-consumer-boundary.mdcCursor rule
- .cursor/rules/caveman-mode.mdcCursor rule
- .cursor/rules/cost-routing.mdcCursor rule
- .cursor/rules/execute-operator-intent.mdcCursor rule
- .cursor/rules/ponytail-lite.mdcCursor rule
- debugSkill
- ponytail-liteSkill
- promptSkill
- pr-reviewSkill
- systematic-debuggingSkill
The file
We split this skill into the rules it gives an agent
7 rules. 0 are common, 0 are distinctive, and 7 are very specific to this project.
0
Common
Most of the 181 edited versions we read keep these.
0
Distinctive
This author's own choices. Worth a look.
7
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
If plan mode is on, stop: say that /ship needs plan mode off (Shift+Tab) and end.
- Very specific: line 35
Do not edit source to make commits cleaner.
- Very specific: line 37
Never stage .env*, private keys, tokens, credential stores, state/, logs, or large generated artifacts.
- Very specific: line 42
Run the project's own checks for the touched areas: package scripts, Makefile, CI workflow, or documented commands, including …
- Very specific: line 48
If origin/main is ahead and you have no local commits, git pull --ff-only first (it refuses to overwrite local changes.
- Very specific: line 54
Never --no-verify, never amend a pushed commit.
- Very specific: line 100
Never delete branches holding user work.
“Common” is measured: kept by at least 6 in 10 of the 181 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
ship
Ship the working tree of the repo in the session's working directory: inspect → understand → split logically → validate → commit → pick strategy → push/PR → verify → report. /ship is not "commit everything and push".
Notes: $ARGUMENTS
Octo root (holds the script): !pwsh -NoProfile -Command '$d = Get-Item "${CLAUDE_SKILL_DIR}"; if ($d.Target) { $d = Get-Item ([string]$d.Target) }; (Resolve-Path (Join-Path $d.FullName "../../..")).Path'
If plan mode is on, stop: say that /ship needs plan mode off (Shift+Tab) and end. Work autonomously; ask only on the stop conditions below.
1. Inspect (read-only)
From the repo root: git status --short --branch, git diff --stat, git diff --cached --stat, git log --oneline -10, git remote -v, then:
pwsh -NoProfile -ExecutionPolicy Bypass -File <octo>/scripts/ship.ps1 -WhatIfIt fetches origin/main and prints the branch, ahead/behind vs origin/main, the origin default branch, commits already ahead, and the delivery mode with its reason (config, GitHub branch protection or rulesets, or unknown). It modifies nothing.
Stop (SHIP BLOCKED) when the state is ambiguous: merge/rebase/cherry-pick in progress, HEAD diverged from origin/main, the origin default branch is not main, or commits already ahead of origin/main that are not clearly part of this work. On another branch with no commits of its own, git switch main is a safe move (it carries the working tree).
2. Split into logical commits
Read the actual diffs (git diff, and the content of untracked files). Group changes into independently understandable commits: one logical change each (feat, fix, refactor, test, docs, chore, build). Never one giant commit because many files changed; never squash logical changes to cut the count; unrelated changes go in separate commits.
- A file with unrelated hunks → partial staging.
git add -pis interactive and unavailable: write the hunks to a patch in the scratchpad andgit apply --cached <patch>, then confirm withgit diff --cached. - Do not edit source to make commits cleaner. Preserve all user work: no
stash,reset --hard,checkout --,restoreorcleanon user files. - Changes outside the session's work that you cannot attribute: leave them uncommitted and list them in the report.
- Never stage
.env*, private keys, tokens, credential stores,state/, logs, or large generated artifacts. A diff containing credential-like values → stop withSHIP BLOCKEDnaming the file and line, never the value.
3. Validate
- Review the final diff of every group.
- Run the project's own checks for the touched areas: package scripts, Makefile, CI workflow, or documented commands, including format/lint/typecheck when they are part of the workflow. Prefer focused checks over full suites; docs-only changes skip checks.
- Failure caused by these changes →
SHIP BLOCKEDbefore committing. Clearly pre-existing and unrelated → report it and continue only if safe. Never fix unrelated failures here. - Repo has
scripts/boundary-audit.ps1→ run it with-Stagedbefore each commit.
4. Commit
If origin/main is ahead and you have no local commits, git pull --ff-only first (it refuses to overwrite local changes; a refusal is a blocker).
For each group: stage exactly its files/hunks, check git diff --cached --stat, commit, then verify with git log -1 --stat and git status --short.
- Message:
type(scope): description, imperative, ≤ 72 chars; neverupdate,changes,fix stuff,misc,work. Optional body: 1-3 bullets on the why. Trailer: the session's commit attribution line, if the system prompt gives one. - Multi-line messages: single-quoted here-string with the closing
'@at column 0, e.g.git commit -m @'...'@. - Never
--no-verify, never amend a pushed commit. A hook failure is fixed at its cause.
5. Deliver
The script ships origin/main..HEAD. It refuses staged-but-uncommitted changes and a HEAD behind origin/main, runs boundary-audit.ps1 (full) when present, and never stashes, rebases or force-pushes. Uncommitted leftovers stay local.
Mode direct (config mode: direct, or main has no protection requiring PR):
pwsh -NoProfile -ExecutionPolicy Bypass -File <octo>/scripts/ship.ps1Pushes HEAD:main (fast-forward only) and checks origin/main equals HEAD. No feature branch, no PR.
Mode pr (config mode: pr, or protection/rulesets that block direct pushes):
pwsh -NoProfile -ExecutionPolicy Bypass -File <octo>/scripts/ship.ps1 -BranchName feature/short-name -PrTitle 'feat(scope): summary' -PrBody @'
## Summary
- What changed
- Why it changed
## Changes
- Important implementation changes
## Validation
- Checks executed and their results (only what actually ran)
## Commits
- <subject of each commit>
<PR attribution line, if the system prompt gives one>
'@- Branch:
feature/<name>for features,fix/<name>for bug fixes,chore/<name>for maintenance; lowercase, named after the actual change. Frommain, the script creates it at HEAD and moves localmainback toorigin/main(the commits stay on the branch). Already on a work branch → that branch is pushed. - PR title: Conventional Commit style. Body: no marketing language; never claim a check passed that did not run.
Mode unknown (protection could not be read): pass the PR arguments anyway. The script tries the direct push and falls back to a PR only when GitHub rejects it for protection. Without PR arguments that rejection exits 3; re-run with them.
6. On failure
- Push rejected as non-fast-forward:
SHIP BLOCKED, commits stay local. Never force-push, never rewrite remote history. ghmissing, auth or permission error:SHIP BLOCKEDwith the exact error.gh pr createfailed: the branch is pushed; report it with the error.- boundary-audit failed: fix the flagged content (amend the unpushed commit that introduced it, or add a commit), re-run step 5. Two failures on the same finding → stop and show the audit output.
- Never delete branches holding user work; never retry with
--forceor--no-verify.
Report
SHIP COMPLETE
Strategy: Direct push | Pull Request
Branch: <branch>
Target: main
Commits:
- <sha> <subject>
Validation:
- <check>: PASS/FAIL/SKIPPED
Remote:
- <push/PR result>
PR:
- <URL, if created>On failure:
SHIP BLOCKED
Reason:
<exact reason>
Changes committed:
<yes/no>
Changes pushed:
<yes/no>
Recommended next action:
<specific action>Also list any uncommitted changes left out on purpose.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

