New Project Developer Walkthrough¶
This walkthrough guides you from zero to your first deployed artifacts in about 15 minutes. You'll install SkillMeat, create a collection, add artifacts from the marketplace, deploy them to your project, and verify everything works.
Prerequisites¶
What you need
- Python 3.9 or later — check with
python --versionorpython3 --version - Basic terminal comfort — you'll run a few CLI commands
- A Claude Code project — a local directory with a
.claude/folder (or you'll create one) - Internet access — to download artifacts from GitHub and the marketplace
- No prior SkillMeat experience required — you're starting fresh
Step 1: Install SkillMeat¶
Choose your preferred installation method:
Verify the installation:
You should see version output like skillmeat, version 0.32.0.
Step 2: Initialize Your Collection¶
A collection is your personal library where you store and manage all your artifacts. Create your first one:
You'll see output like:
Your collection is now ready to receive artifacts. You can explore it if curious:
Step 3: Add Your First Artifact from the Marketplace¶
Now you'll add a real artifact. Let's start with a useful skill from Anthropic's official marketplace:
SkillMeat will show you a security prompt. Read it to understand what you're adding, then confirm:
Fetching artifact from GitHub...
Added skill: canvas
Location: ~/.skillmeat/collection/skills/canvas/
What just happened: SkillMeat downloaded the Canvas Design skill from GitHub, verified it, and stored it in your collection.
Explore Your Collection¶
See what's in your collection now:
You should see:
Artifacts (1)
┌────────┬────────┬─────────┐
│ Name │ Type │ Origin │
├────────┼────────┼─────────┤
│ canvas │ skill │ github │
└────────┴────────┴─────────┘
Step 4: (Optional) Add More Artifacts¶
Feel free to experiment with additional artifacts from the marketplace:
# Add a different skill
skillmeat add skill anthropics/skills/document
# Or add a command
skillmeat add command user/repo/commands/review
List your collection again to see them:
Step 5: Deploy to Your Project¶
Now you'll deploy your artifacts to your Claude Code project (the .claude/ directory). First, navigate to your project:
If you don't have a .claude/ directory yet, SkillMeat will create it for you.
Deploy the canvas skill:
You'll see output like:
Check what was deployed:
You should see canvas/ directory with the skill files inside.
If you added more artifacts, deploy them too:
Step 6: Verify in the Web UI (Optional but Recommended)¶
The web UI gives you a visual overview of your collection and deployed artifacts. Start it:
Wait for both servers to start. You'll see:
Using the Web UI¶
- Open http://localhost:3000 in your browser
- Click Collection in the left sidebar
- You'll see all your artifacts listed (canvas, and any others you added)
- Click any artifact to see details: version, metadata, what it does
- Click Projects in the left sidebar
- Select your current project
- Click Deployed tab to see all artifacts deployed to this project
- Each shows deployment status, version, and when it was deployed
- [ ] Your artifacts appear in Collection view
- [ ] Your artifacts appear in Projects → [Your Project] → Deployed tab
- [ ] Artifact versions are visible (e.g., "v1.2.0" or "latest")
- [ ] Timestamps show "just now" or recent times
To stop the web UI, press Ctrl+C in the terminal.
Step 7: Quick Verification (CLI Only)¶
If you prefer not to use the web UI, verify everything via the CLI:
# Check your collection
skillmeat list
# Check what's deployed in your project
cd /path/to/your/project
skillmeat status
# View detailed info on a deployed artifact
skillmeat show canvas
Troubleshooting¶
"skillmeat: command not found"¶
You need to activate your virtual environment or reinstall:
# If using virtual environment
source ~/.skillmeat-venv/bin/activate
# Then verify
skillmeat --version
"Collection already exists"¶
You've already run skillmeat init. To use the existing collection:
Or to create a new named collection:
GitHub rate limit errors¶
If you see "rate limit exceeded," set a GitHub token to increase limits:
Get a token from GitHub Settings → Developer settings → Personal access tokens (classic). You don't need special permissions for public repos.
Artifact fails to deploy¶
Check that your project directory has a .claude/ folder:
If not, create it:
Then retry:
What Just Happened¶
You've successfully:
- ✅ Installed SkillMeat
- ✅ Created a personal collection
- ✅ Added artifacts from the marketplace
- ✅ Deployed them to your project
- ✅ Verified everything works (via CLI and optionally web UI)
Your artifacts are now available in your Claude Code projects. Use them in your .claude/ directory just like any other Claude Code artifacts.
Next Steps¶
Now that you have your first artifacts deployed, explore deeper workflows:
- See Also (coming in v1 docs rollout):
docs/user/guides/collection-init-config.md— Understand collection configuration and settingsdocs/user/guides/adding-artifacts.md— Learn all ways to add artifacts (marketplace, GitHub, local files)-
docs/user/guides/deploying-artifacts.md— Master deployment options, including Git PR workflows for teams -
Quick links to common tasks:
- Discovery & Import Quick Guide — Scan your existing
.claude/directory for artifacts to import - Marketplace Usage Guide — Browse and search the marketplace for more artifacts
- Web UI Guide — Get more comfortable with the visual interface
-
CLI Reference — Full list of all available commands
-
Ready for more?
- Do you have an existing project with artifacts? See the Existing Project Migrant Walkthrough
- Setting up for a team? See the Team / Enterprise Admin Walkthrough (coming soon)
- Using Backstage? See the Backstage Platform Engineer Walkthrough (coming soon)
Getting Help¶
- View help for any command:
skillmeat --helporskillmeat deploy --help - Check the CLI Reference for detailed command documentation
- Review Troubleshooting Guide for common issues
- Ask questions in the SkillMeat community or file an issue on GitHub
Time to complete: ~15 minutes
Complexity: Beginner
Prerequisites met: ✅