Product · Ship

Closed beta

The shipping floor for AI‑native engineering teams.

Ship runs the loop from ticket to merged PR. Eight named states, fifteen agent specialists, scheduled routines that fire whether or not the demo is watching. Repo, tracker, policy, knowledge, and evidence — one workspace, one accountable loop. Built by the team that wrote the book on it.

Pair with Lighthouse to keep the memory across every run →

Ship ships Ship

The team that built Ship runs its delivery on Ship.

Every production line in this repository was authored by an AI executor working inside the same workspace this page describes. The four numbers below are the receipts.

236k+

Lines in production

0

Hand-typed by humans

608

Commits in 30 days

7

Team, none typing code

Live cadence · last 30 days

DORA‑4 as a load‑bearing metric, not a bumper sticker.

The four numbers SRE‑literate buyers actually read. Ours come from Ship's own operations, recomputed every render. Two Elite, two High — on a stack where every commit was AI‑authored.

7.3/day

Deployment frequency

Elite

5m

Lead time for changes

Elite

13%

Change failure rate

High

1.5h

Mean time to recovery

High
Ship Analytics dashboard with DORA-4 charts and live system status

Live dashboard from Ship's own analytics surface.

What Ship is

A loop with named owners, fenced agents, and a paper trail.

Six properties that compose into one operating model. None of them are features in isolation; together they're what makes the loop legible to a CTO at a glance.

The loop

Eight named states from ticket to merged PR

Backlog → Planning → Executing → Reviewing → Awaiting input → Blocked → Closed, plus a Default flow. Every ticket is in exactly one. A new hire reads the board on day one.

Named owners

Specialists, not roles in spirit

Planning bundle, Developer, Validation bundle, Reviewer, DevOps, Designer, Clarification specialist. Each stage names which agent acts and which artifact it produces.

Routines

Scheduled jobs that fire whether the demo is watching or not

Self-heal, daily digest, weekly audit, retro. Picked up the moment a ticket is ready — not on a wall clock the SDLC can't trust.

Evidence

Audit trail as a side-effect, not a project

Every decision, retrieval, and outcome captured as an artifact. The Audit log filters by surface and time. Compliance is grep, not a meeting.

Knowledge

Curated facts the agents actually read

Buckets per area — Security, Inbox, Shipctl, Authentication, Configuration. Sources auto-refresh on cadence. Nothing publishes silently — every fact passes through review.

Owner-first

Humans stay accountable

Decisions, approvals, clarifications, retros land in the Inbox. Disposition (Address / Dismiss / Escalate) is a record, not a button-click. The merge is yours.

The surfaces, in order

Open the console. Drain the Inbox. Walk the pipeline.

Four surfaces carry every operator's day. The same four — in the same order — every morning.

Workspace home

Four priority buckets and the cadence line

The home view is four columns: Needs you, Active, Drafts, Parked. Every card is a ticket, a decision, or an autopsy. The footer shows the most recent merged PR with timestamp — if you see nothing there, the system is quietly doing nothing. That's a finding, not a feature.

Read the docs
Four priority buckets and the cadence line

Inbox

Decisions queued, not buried

Only items that need a human land here: stuck PRs (no activity 24h+), orphan tickets skipped at a stage, learning captures from yesterday's retro. Address, dismiss, or escalate each — every disposition is logged.

Read the docs
Decisions queued, not buried

Process · SDLC

The seven-stage pipeline with named specialists

One ticket walks all eight states in order. Each stage names which agent acts there: Implementation specialist, Validation bundle, Code reviewer. You can stand a new hire in front of the board on day one and they can read it.

Read the docs
The seven-stage pipeline with named specialists

Audit

Fifty rows is one filter away

Every workspace, member, integration, pipeline, repo, improvement, clarification, invite, and agent action — captured. Filter by surface and time. If a question can be answered by grep, it is answered here.

Read the docs
Fifty rows is one filter away

The specialists

Named agents. Versioned roles. Swappable executors.

A specialist is a role definition — "developer", "code reviewer", "validation bundle" — pinned to a stage. The executor (Cursor, Claude Code, Codex, Copilot) plugs in underneath. Swap executors tomorrow and the routine, the specialist, and the process stay unchanged. That separation is the whole product.

Pipeline owners

Planning bundleDeveloperValidation bundleReviewer

Platform

DevOpsSecurityQA automation

Intake / clarify

Intake specialistClarification specialistDesigner review

Closed beta

Open the console and walk the loop. Or read the book first.

We onboard a small cohort at a time. If you run a product line and want a delivery loop your team can read at a glance — start here.