How long does the integration take?

Change one base URL in an OpenAI-compatible client, or install the Python or TypeScript SDK; the quickstart routes a first request in a few minutes.

Will this lock us into another vendor?

No. The router is Apache-2.0 open source, your provider keys stay yours, and the API is OpenAI-compatible, so leaving is a change of base URL.

What happens when a provider fails?

The router falls back to the next model your policy allows and records the fallback in the trace, so an outage costs a retry, not an incident.

ship - Prompt - OpenSmartRoute

Name: ship
Author: KhourySpecialProjects

Prompt file imported from KhourySpecialProjects/odyssey (.claude/commands/ship.md). Fill in {{arguments}} before use. Copyright stays with the author.

Prepare Branch for PR

Run pre-ship checks and prepare a PR-ready summary for the current branch.

Mode: {{arguments}}

If no mode specified, run full checks.

Steps

1. Verify clean state

Run git status to check for uncommitted changes. If there are unstaged changes, warn the human and ask whether to proceed or stop.

2. Run full CI

Run the same checks as /ci full:

bash scripts/ci-local.sh

If CI fails, show the failures and STOP. Do not proceed to PR preparation with failing checks. Tell the human what needs fixing.

3. Gather branch context

git log --oneline production..HEAD
git diff --stat production..HEAD

Count files changed, insertions, deletions. Identify which areas were touched (frontend components, request functions, Strapi schemas, tests, config).

4. Check for plan/spec

Look in docs/plans/ for the active plan and spec files. Extract:

The problem statement from the spec
Acceptance criteria and their completion status from the plan
Any audit verdict if an audit was run

5. Generate PR summary

Present a ready-to-copy PR description:

## Summary

- [1-3 bullet points from the spec/plan]

## Changes

- [Categorized file changes: frontend, backend, tests, config]

## Acceptance Criteria

- [x] [Each criterion from the spec, checked if met]

## Test Plan

- [ ] CI passes locally (Prettier, ESLint, Jest, builds)
- [ ] [Additional test steps based on what changed]

🤖 Generated with [Claude Code](https://claude.com/claude-code)

6. Final status

Ship Check
══════════
CI: ✓ All checks passed / ✗ [failures]
Branch: [branch name] — [N] commits ahead of production
Changes: [N] files changed (+[insertions] -[deletions])
Plan: [plan file name] — [N/M] tasks complete
Audit: [verdict if exists, or "not audited"]

Ready to push: YES / NO — [reason if no]

If everything passes, tell the human: "Branch is ready. Push when you're ready." Do NOT push — the human pushes manually per project policy.

Install ship and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

# Prepare Branch for PR Run pre-ship checks and prepare a PR-ready summary for the current branch. ## Mode: {{arguments}} If no mode specified, run full checks. ## Steps ### 1. Verify clean state Run `git status` to check for uncommitted changes. If there are unstaged changes, warn the human and ask whether to proceed or stop. ### 2. Run full CI Run the same checks as `/ci full`: ```bash bash scripts/ci-local.sh ``` If CI fails, show the failures and STOP. Do not proceed to PR preparation with failing checks. Tell the human what needs fixing. ### 3. Gather branch context ```bash git log --oneline production..HEAD git diff --stat production..HEAD ``` Count files changed, insertions, deletions. Identify which areas were touched (frontend components, request functions, Strapi schemas, tests, config). ### 4. Check for plan/spec Look in `docs/plans/` for the active plan and spec files. Extract: - The problem statement from the spec - Acceptance criteria and their completion status from the plan - Any audit verdict if an audit was run ### 5. Generate PR summary Present a ready-to-copy PR description: ```markdown ## Summary - [1-3 bullet points from the spec/plan] ## Changes - [Categorized file changes: frontend, backend, tests, config] ## Acceptance Criteria - [x] [Each criterion from the spec, checked if met] ## Test Plan - [ ] CI passes locally (Prettier, ESLint, Jest, builds) - [ ] [Additional test steps based on what changed] 🤖 Generated with [Claude Code](https://claude.com/claude-code) ``` ### 6. Final status ``` Ship Check ══════════ CI: ✓ All checks passed / ✗ [failures] Branch: [branch name] — [N] commits ahead of production Changes: [N] files changed (+[insertions] -[deletions]) Plan: [plan file name] — [N/M] tasks complete Audit: [verdict if exists, or "not audited"] Ready to push: YES / NO — [reason if no] ``` If everything passes, tell the human: "Branch is ready. Push when you're ready." Do NOT push — the human pushes manually per project policy.

{ "prompt": "# Prepare Branch for PR\n\nRun pre-ship checks and prepare a PR-ready summary for the current branch.\n\n## Mode: {{arguments}}\n\nIf no mode specified, run full checks.\n\n## Steps\n\n### 1. Verify clean state\n\nRun `git status` to check for uncommitted changes. If there are unstaged changes,\nwarn the human and ask whether to proceed or stop.\n\n### 2. Run full CI\n\nRun the same checks as `/ci full`:\n\n```bash\nbash scripts/ci-local.sh\n```\n\nIf CI fails, show the failures and STOP. Do not proceed to PR preparation\nwith failing checks. Tell the human what needs fixing.\n\n### 3. Gather branch context\n\n```bash\ngit log --oneline production..HEAD\ngit diff --stat production..HEAD\n```\n\nCount files changed, insertions, deletions. Identify which areas were touched\n(frontend components, request functions, Strapi schemas, tests, config).\n\n### 4. Check for plan/spec\n\nLook in `docs/plans/` for the active plan and spec files. Extract:\n\n- The problem statement from the spec\n- Acceptance criteria and their completion status from the plan\n- Any audit verdict if an audit was run\n\n### 5. Generate PR summary\n\nPresent a ready-to-copy PR description:\n\n```markdown\n## Summary\n\n- [1-3 bullet points from the spec/plan]\n\n## Changes\n\n- [Categorized file changes: frontend, backend, tests, config]\n\n## Acceptance Criteria\n\n- [x] [Each criterion from the spec, checked if met]\n\n## Test Plan\n\n- [ ] CI passes locally (Prettier, ESLint, Jest, builds)\n- [ ] [Additional test steps based on what changed]\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n```\n\n### 6. Final status\n\n```\nShip Check\n══════════\nCI: ✓ All checks passed / ✗ [failures]\nBranch: [branch name] — [N] commits ahead of production\nChanges: [N] files changed (+[insertions] -[deletions])\nPlan: [plan file name] — [N/M] tasks complete\nAudit: [verdict if exists, or \"not audited\"]\n\nReady to push: YES / NO — [reason if no]\n```\n\nIf everything passes, tell the human: \"Branch is ready. Push when you're ready.\"\nDo NOT push — the human pushes manually per project policy.", "variables": [ "arguments" ], "notes": "Run pre-ship checks and prepare a PR-ready summary for the current branch.", "metadata": { "source": { "provider": "github-claude-commands", "repository": "https://github.com/KhourySpecialProjects/odyssey", "path": ".claude/commands/ship.md", "ref": "2fa7a7f4e8136a5735af659871f37cd5e78e91e2", "url": "https://github.com/KhourySpecialProjects/odyssey/blob/2fa7a7f4e8136a5735af659871f37cd5e78e91e2/.claude/commands/ship.md", "key": "KhourySpecialProjects/odyssey/.claude/commands/ship.md" } } }

ship

About

Prepare Branch for PR

Mode: {{arguments}}

Steps

1. Verify clean state

2. Run full CI

3. Gather branch context

4. Check for plan/spec

5. Generate PR summary

6. Final status

Install ship and it becomes one more candidate the router can pick - when it fits.

What it costs to adopt - in engineering time.

One base URL

No vendor lock-in

Failover built in

Use it

Manifest

Reviews