ship

A command that takes a code change through quality checks, commits, pushes, opens a GitHub pull request, watches CI, reviews it, and merges it when allowed. CI is automated checking that runs against the change.

In plain words
What is it for?
Running the full test and QA suite, recording each result, creating and monitoring a pull request, and merging only when all required gates pass.
Why use it?
It organizes the release process and prevents merging when required checks fail, time out, or cannot be verified.

Command

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 commands/alphacornutum/layercake/ship
Clone the repo
git clone --depth 1 https://github.com/alphacornutum/layercake
Per session 27 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,067 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 $0.00027 $0.01067
Opus 5 $0.00014 $0.00534
Sonnet 5 $0.00005 $0.00213
Haiku 4.5 $0.00003 $0.00107

Measured 2d ago against content hash a2308a735b3a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

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 2d ago.

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.

.ai/src/commands/ship.md · 76 lines

How it starts

The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Get the current work release-ready end-to-end$ARGUMENTS.

Quality gates (hard stop on merge)

Every command in the full QA block below is a required quality gate, including npm run test:ae.

  • Run the full suite to completion. If one gate fails or times out, continue the remaining gates/tasks so the final report lists everything that failed — not only the first problem.
  • Record pass/fail/skip/timeout for each gate as you go. Do not abort the QA block early just because one step failed (unless a later step is impossible without an earlier artifact, e.g. build after a broken install — note that dependency and still run every independent gate you can).
  • If any gate fails, times out, is skipped because the environment is broken, or cannot be verified: do not merge.
  • At the end (and on the PR if one exists), post a comprehensive status: every gate with outcome + evidence for failures/timeouts, plus what must be fixed or re-run.
  • Treat “CI is green” as necessary but not sufficient when local gates (especially test:ae) did not pass. GitHub CI does not run After Effects host tests — local test:ae is still mandatory before merge when the host platform is macOS/Windows.
  • Do not reinterpret hangs, recovery-dialog flakes, or “it passed earlier in the session” as a pass for this ship run.

Exception: On Linux (or when test:ae correctly no-ops via describe.skipIf(!hasHost) because no AE host is configured), document that skip explicitly in the PR and treat it as a clean skip — not a failed/hung run.

1. Prepare and full QA

  • Finish any incomplete OpenSpec/docs/agent work needed for a coherent release.
  • If you edited .ai/src/, run agentsync sync first.
  • Make sure docs are up to date:
npm run docs:fetch
npm run docs:allowlist
  • Run full QA. Prefer running gates so a single failure does not hide later ones (e.g. chain with ; / record exit codes, or run steps sequentially and continue after failures). In Cursor, use Shell required_permissions: ["all"]agentsync check always fails in the sandbox (Failed to prepare temporary workspace for check); do not attempt a sandboxed first try.

Read the full file on GitHub · 76 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. 2d ago First seen · 76 lines · 27 tokens per session scan A a2308a735b3a

Subscribe to this mod's changes

ship is a command published in the GitHub repository alphacornutum/layercake (0 stars, last pushed 1mo ago), licensed MIT. It adds 27 tokens to every session and 1,067 once invoked, about $0.0001 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.