shipkit-ship

A build workflow that takes a written feature plan and carries out its steps until the defined completion criteria are met.

In plain words
What is it for?
Use it after planning a feature when you want the coding agent to build, integrate, and verify the planned work.
Why use it?
It removes the need to manually coordinate parallel implementation work, combine the results, and check whether the feature is truly finished.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/stefan-stepzero/shipkit/shipkit-ship
Any agent
npx skills add stefan-stepzero/shipkit --skill shipkit-ship
Clone the repo
git clone --depth 1 https://github.com/stefan-stepzero/shipkit

Made for: Claude Code, Codex.

Per session 63 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,127 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00063 $0.01127
Opus 5 $0.00032 $0.00563
Sonnet 5 $0.00013 $0.00225
Haiku 4.5 $0.00006 $0.00113

Measured yesterday against content hash 44498f2acc92, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

shipkit-ship scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

install/skills/shipkit-ship/SKILL.md · 103 lines

How it starts

The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.

shipkit-ship — Build to Done

Purpose: The consumer-facing BUILD entry point. Takes a plan (.shipkit/plans/) and drives it to a confirmed, ground-truth completion bar.

shipkit-ship is a thin caller. It does not implement its own build loop — it states the what (steps + bar) and hands off to the engine, which owns the how (partition → fan-out → reconcile → integrate → stop-at-bar). This is Shipkit-native drive-to-done: one engine, one completion contract.

Read first: install/shared/references/core-automation.md — the contract the engine runs. This skill only assembles the call.


What it does

Invoke shipkit-orchestrate in autonomous mode over the plan. Nothing more — the engine runs inline in the main session, launches the parallel build, reconciles against ground truth, integrates serially, and stops at the bar.

Pass the engine the four inputs (calling interface in install/skills/shipkit-orchestrate/SKILL.md):

Input For a build
steps the plan's phases / file-ownership clusters, in dependency order — the units to build (from .shipkit/plans/<feature>.json)
bar the ground-truth completion criteria (below) — stated explicitly; the engine advances only when they are true on disk / in git
mode autonomous (build runs unattended to the bar) — add the scale toggle for very large / long builds
elicitive normally none — a build should run with the spec/plan already resolved. If a step genuinely needs a human decision, mark it elicitive so the engine bubbles it (never inside a workflow), rather than letting a worker guess.

The bar (state it — don't let a worker self-report "done")

A build is done only when all of these are reconciled (verified, not reported):

  1. Functional surface complete — every element the no-gaps spec named across the four dimensions (applications / datastores / contracts / integrations) is built, not just the happy path. A frontend that implies a backend has the backend.
  2. Plan acceptance met — every acceptance criterion / phase gate in the plan passes.
  3. Tests green — the plan's named checks (tests, build, lint) pass.
  4. Integrated — work is committed, rebased onto the target branch, and landed per the integration policy (serial, one lane at a time, never forced past unverified work).

Read the full file on GitHub · 103 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 103 lines · 63 tokens per session scan A 44498f2acc92

Subscribe to this mod's changes

shipkit-ship is a skill published in the GitHub repository stefan-stepzero/shipkit (1 stars, last pushed 1mo ago), licensed MIT. It adds 63 tokens to every session and 1,127 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.