Design Ship logo

Design Ship

Community
yonatangross
design-ship

One-shot pipeline turning a claude.ai/design link into a pull request: scaffold via /ork:design-import, stories and specs via /ork:cover, browser verification via /ork:expect, then open the PR. Use when a design link should come back as a PR with no intermediate steps; if all you need is the components written to disk, run /ork:design-import instead.

Overview

Publisheryonatangross
Repositoryorchestkit
Skill namedesign-ship
Stars
298
Forks
35
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by yonatangross on GitHub. Read the source before you install it.

Installation

Install the Design Ship AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/yonatangross/orchestkit/tree/main/plugins/ork/skills/design-ship
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/yonatangross/orchestkit.git /tmp/orchestkit
mkdir -p .claude/skills
cp -r /tmp/orchestkit/plugins/ork/skills/design-ship .claude/skills/design-ship
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Design Ship in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Design Ship on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Design Ship is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Design Ship

One command: Claude Design handoff URL → reviewable GitHub PR.

bash
design-ship https://claude.ai/design/abc123     # From handoff URL
design-ship /tmp/handoff-bundle.json            # From local file

When to use

You have a Claude Design handoff URL (or file) and want a PR opened against the current branch. No intermediate steps, no manual test generation, no manual PR drafting.

For just-the-import (no tests, no PR), use design-import instead.

Pipeline

Handoff bundle (URL or file)
  │
  ▼
┌──────────────────────────────┐
│ 1. design-import         │  Scaffold components, write provenance
│    (delegates to orchestrator │  Tokens reconciled
│     for parse + dedup)        │  Components written or reused
└──────────┬───────────────────┘
           │
           ▼
┌──────────────────────────────┐
│ 2. cover                 │  Storybook stories per new component
│    (Storybook + Playwright)   │  Playwright E2E for new routes/views
└──────────┬───────────────────┘
           │
           ▼
┌──────────────────────────────┐
│ 3. expect                │  Diff-aware browser verification
│    (CDP + ARIA-tree-first)    │  Screenshots saved for PR body
└──────────┬───────────────────┘
           │
           ▼
┌──────────────────────────────┐
│ 4. create-pr             │  PR opened with:
│                               │   - Claude Design URL link
│                               │   - Before/after screenshots
│                               │   - Component scaffold list
│                               │   - Coverage delta
│                               │   - Label: claude-design
└──────────┬───────────────────┘
           │
           ▼
┌──────────────────────────────┐
│ 5. UPDATE PROVENANCE          │  Patch .claude/design-handoffs/<id>.json
│                               │  with the opened PR number + URL
└──────────────────────────────┘

Pre-flight

Before starting the pipeline, verify the working tree is clean enough to ship:

python
status = Bash("git status --porcelain")
if status:
    AskUserQuestion(questions=[{
      "question": "Working tree has uncommitted changes. How to proceed?",
      "header": "Pre-flight",
      "options": [
        {"label": "Commit them first", "description": "Run commit, then proceed with ship"},
        {"label": "Stash and restore", "description": "Stash now, restore after PR opens"},
        {"label": "Include in this PR", "description": "Risky — only if changes are related"},
        {"label": "Cancel", "description": "Abort design-ship"}
      ],
      "multiSelect": False
    }])

current_branch = Bash("git branch --show-current")
if current_branch in ("main", "master", "dev"):
    AskUserQuestion(questions=[{
      "question": f"You're on {current_branch}. Create a feature branch?",
      "header": "Branch",
      "options": [
        {"label": "Yes — feat/design-ship-<bundle-id>", "description": "Recommended"},
        {"label": "Cancel", "description": "Switch branch yourself first"}
      ],
      "multiSelect": False
    }])

Phase 1 — Import

Delegate to design-import with the bundle argument. Capture its output (the import manifest) — it includes the bundle_id, list of components written, and the provenance file path.

If import fails, stop. Do NOT proceed to test generation against partially-imported components.

Phase 2 — Cover

Run cover scoped to the components written by phase 1:

python
component_paths = [c["path"] for c in import_result["components"] if c["decision"] != "reuse"]

Agent(
  subagent_type="ork:test-generator",
  description="Cover the freshly-imported components",
  prompt=f"""Use the cover skill to generate tests for these components (scoped — do not regenerate suites for unrelated files):
  {component_paths}

  Tiers needed:
  - Storybook stories per component (mandatory)
  - Playwright E2E if any new route was registered (check src/app or src/pages)
  - Unit tests if the components include non-trivial logic (skip for pure-presentation)
  """
)

Phase 3 — Expect

Run expect to do diff-aware browser verification. This produces the screenshots that go into the PR body:

python
Agent(
  subagent_type="ork:expect-agent",
  description="Diff-aware verification of imported components",
  prompt=f"""Use the expect skill on the current diff.

  Extra: capture before/after screenshots for each new component
  and save them to .claude/design-handoffs/{bundle_id}/screenshots/.
  These will be uploaded to the PR body.

  If verification fails, do not block — report findings and continue.
  Failures will surface in the PR review.
  """
)

Phase 4 — Create PR

Use create-pr with a body templated from the bundle and verification results:

python
pr_body = f"""## Summary
Generated from Claude Design handoff: {bundle_url}

Bundle ID: `{bundle_id}` · Provenance: `.claude/design-handoffs/{bundle_id}.json`

## Components
| Decision | Component | Path |
|---|---|---|
{table_from(import_result['components'])}

## Tokens
- Added: {len(token_diff['added'])}
- Modified: {len(token_diff['modified'])}
- Conflicts (resolved): {len(token_diff['conflicts'])}

## Verification (`expect`)
{expect_summary}

## Coverage delta (`cover`)
{coverage_delta}

## Screenshots
{screenshot_table}

---
🤖 Opened by `design-ship` · Closes: (none — link issues manually)
"""

Bash(f"gh pr create --title '{pr_title}' --body @- <<< '{pr_body}' --label claude-design")

Phase 5 — Update provenance

Patch the provenance file with the opened PR number and URL:

python
provenance = Read(f".claude/design-handoffs/{bundle_id}.json")
provenance["pr"] = {"number": pr_number, "url": pr_url, "opened_at": now()}
Write(f".claude/design-handoffs/{bundle_id}.json", provenance)

Failure modes

PhaseFailureBehavior
Pre-flightDirty tree, on mainAskUserQuestion, do not proceed silently
1. ImportSchema deviation, token conflictdesign-import already prompts; if user cancels, abort
2. CoverTest generation failsContinue to expect — tests are recoverable in PR review
3. ExpectBrowser verification failsContinue to PR — failures noted in PR body
4. Create PRgh not authenticatedStop, surface auth error
5. ProvenanceFile not writableWarn but do not fail (PR is already open)

Composition

SkillRole
design-importPhase 1 — scaffold components
coverPhase 2 — Storybook + Playwright
expectPhase 3 — diff-aware browser verification
create-prPhase 4 — open PR with templated body
remember / memoryProvenance updates

NOT this skill's job

ConcernOwned by
Parsing bundlesclaude-design-orchestrator agent (called by design-import)
Component dedupOrchestrator agent (called by design-import)
Test execution (CI)GitHub Actions — this skill only generates tests
Merging the PRHuman reviewer — never auto-merge

Limitations

  • No auto-merge: this skill opens the PR; humans review and merge.
  • No iteration loop: if the PR review surfaces design changes, you currently re-export from Claude Design and re-run design-ship. The reverse-path (live UI drift → refinement prompt for claude.ai/design) is tracked in M124 — Claude Design Drift Sync (Bet B): #1391 (PR hook), #1393 (design-iterate), #1395 (design-drift-detector agent).
  • Single-bundle scope: one invocation, one bundle, one PR. Multi-bundle batches are not supported.

Frequently asked questions

What does the Design Ship AI skill do?

One-shot pipeline turning a claude.ai/design link into a pull request: scaffold via /ork:design-import, stories and specs via /ork:cover, browser verification via /ork:expect, then open the PR. Use when a design link should come back as a PR with no intermediate steps; if all you need is the components written to disk, run /ork:design-import instead.

Why use Design Ship on TypingMind?

Because you install it once and use it with any model. Design Ship is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Design Ship in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/yonatangross/orchestkit/tree/main/plugins/ork/skills/design-ship. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Design Ship?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Design Ship?

As many as you like. As long as a model supports skills, you can use Design Ship with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Design Ship AI skill free?

Yes. It is published on GitHub by yonatangross under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇