Scan & Import Existing Artifacts¶
Found artifacts scattered across your .claude/ directories? SkillMeat can automatically discover and import them into your collection in one workflow. This guide covers scanning, importing, organizing, and verifying artifacts.
Walkthrough available
See the Existing Project Migrant Walkthrough for a complete migration journey.
When to use this guide
Use this guide when you have an existing .claude/ directory with skills, commands, agents, hooks, or MCP servers that you want to bring into your SkillMeat collection.
For step-by-step scanning and import mechanics, see Discovery & Import Quick Guide.
Why Scan & Import?¶
Discovery saves time and reduces manual work:
| Scenario | Benefit |
|---|---|
| Artifact Audit | Understand what artifacts you've created locally and their current state |
| Migration | Move all local artifacts from .claude/ into your managed collection |
| Deduplication | Identify duplicate artifacts and consolidate them |
| Organization | Establish a single source of truth with tags, metadata, and aliases |
Prerequisites¶
Before scanning, ensure:
- SkillMeat initialized — Your collection exists at
~/.skillmeat/collection/ - If not: Run
skillmeat init(see Quickstart) - Target
.claude/directory identified — The project directory containing artifacts to scan - Web UI running (for Web UI workflow) — Optional but recommended for visual organization
- Start with:
skillmeat web dev
Workflow Overview¶
The complete scan-to-organization workflow has 5 steps:
- Scan for artifacts — Discover what exists
- Review results — Understand what was found
- Import artifacts — Add to your collection
- Organize & tag — Add metadata for future discovery
- Verify & deploy — Confirm success and deploy to projects
Scan for Artifacts¶
CLI Scan¶
Scan a specific directory for artifacts:
# Scan current project directory
skillmeat scan .
# Scan specific path
skillmeat scan /path/to/project
# Scan and show detailed output
skillmeat scan . --verbose
Expected output:
- Navigate to Artifacts page
- Click Discover button (top of artifact list)
- The system auto-scans your project's
.claude/directory - Review artifacts in the modal that appears
Understanding Scan Results¶
The scan returns artifacts grouped by type with status:
| Field | Meaning |
|---|---|
| Status: New | Artifact not in collection; ready to import |
| Status: Exists | Artifact already in collection (can skip or re-import) |
| Status: Conflict | Same artifact name exists in collection with different content |
| Status: Error | Artifact missing required metadata files (fix manually or skip) |
Common scan results
- Most artifacts show
Status: New— they're fresh, ready to import - Re-running scan after import shows
Status: Exists— normal, expected - Conflicts are rare; they indicate duplicate names with different content (decide which to keep)
Review Scan Results¶
Via CLI¶
List scan results with filters:
The Discovery Modal shows all results with:
- Artifact name and type (skill, command, agent, etc.)
- Description (from metadata)
- Tags (if present in artifact)
- Status badge (New, Exists, Conflict, Error)
- Checkbox for selective import
Filtering in modal: - Click Type dropdown to filter by artifact type - Use Search box to find specific artifacts by name - Click Status to focus on New, Exists, or Conflict
Review Checklist¶
Before importing, scan the results for:
- [ ] Artifacts you recognize — Most scanned artifacts should be yours
- [ ] Unexpected artifacts — Any files that don't look like artifacts? Check them manually
- [ ] Duplicates — Same artifact name in multiple locations? (rare, but possible)
- [ ] Errors — Any artifacts with
Status: Error? (usually missing metadata, safe to skip)
Resolving errors
If an artifact shows Status: Error, it's missing required files like SKILL.md or COMMAND.md.
Options: - Fix it manually — Add the missing metadata file, then re-scan - Skip it — Proceed without importing; you can import it later - Investigate — Check if the directory really is an artifact or just a regular folder
Import Artifacts¶
Import All Found Artifacts¶
- In the Discovery Modal, all artifacts are pre-selected by default
- Click Import All (or Confirm Import)
- Progress bar shows import status
- Success message lists imported artifacts
Import Selected Artifacts¶
- In the Discovery Modal, uncheck artifacts you don't want
- Check only the artifacts you want to import
- Click Import Selected
- Progress bar shows which artifacts succeeded
Import is deployable
When you import an artifact from your own .claude/ directory, SkillMeat:
- Adds it to your collection
- Automatically deploys it back to the project
- This means no separate deploy step is needed — it's ready to use immediately
Handling Import Conflicts¶
If an artifact exists with the same name but different content:
- Look for artifacts with Status: Conflict badge
- Click the conflict badge to see details:
- What's in your collection now
- What's being scanned
- Choose:
- Keep Collection Version — Skip this import
- Replace with Scanned Version — Use the scanned artifact
- Proceed with import
Resolving conflicts
Conflicts usually mean you've updated an artifact since last import. Review both versions in the modal and decide which is more current.
Post-Import Organization¶
After importing, organize your artifacts with tags, aliases, and groupings for easy discovery later.
Tagging Imported Artifacts¶
- Navigate to Artifacts
- Search for or scroll to the imported artifact
- Click the artifact to open details
- Click Tags section
- Add tags separated by commas (e.g.,
imported, python, productivity) - Save
Using Aliases for Quick Access¶
- Open artifact details (search for artifact, click it)
- Look for Aliases field
- Enter comma-separated short names (e.g.,
skill, my-skill) - Save
Grouping into Bundles (Optional)¶
Bundles group related artifacts for collective deployment:
- Navigate to Collections or Bundles section
- Click New Bundle
- Name:
My Toolkit| Description:Collection of my tools - Click Create
- Click Add Artifacts
- Search for and select the imported artifacts
- Save
When to use bundles
Bundles are useful when: - You have 3+ related artifacts that are always deployed together - You want to share a curated set with teammates - You need version control for a cohesive toolset
For most users, tags are sufficient for organization. See Bundle & Composite Authoring Guide for full bundle workflow.
Updating Metadata¶
- Open artifact details
- Click Edit or pencil icon
- Update fields:
- Description
- Version
- Owner/Scope
- Tags
- Visibility
- Save
Verification¶
Verify imports succeeded and are ready to use:
CLI Verification¶
- Navigate to Artifacts page
- Filter by Source: Imported (if available)
- Search for imported artifacts by name or tag
- Click each artifact to verify:
- Description is present
- Tags are applied
- Metadata is complete
Deployment Verification¶
Imported artifacts are automatically deployed to the project. Verify they exist:
Verification checklist
- [ ] Artifacts appear in
skillmeat list - [ ] Tags/metadata are applied
- [ ] Artifacts are deployed to project (in
.claude/directory) - [ ]
skillmeat search <name>finds them - [ ] Running
skillmeat statusshows them as deployed
Troubleshooting¶
Scan finds no artifacts¶
Problem: Directory scanned but no artifacts detected.
Causes & solutions:
- Wrong directory — Scan the project directory, not .claude/ itself
SKILL.md, COMMAND.md, etc.
- Check directory for these files: ls -la .claude/skills/
- Nested directories not recognized — If artifacts are in custom subdirectories
- Manually import using skillmeat add instead (see Adding Artifacts)
Import fails with "conflict"¶
Problem: Artifact import blocked due to name conflict.
Solution:
- Review what's in your collection: skillmeat info <artifact-name>
- Choose to keep or replace version (use --force flag to auto-replace)
- Or rename the scanned artifact before importing
Tags don't persist¶
Problem: Tags added after import disappear on refresh.
Solution:
- Tags must be added in your collection, not in source files
- After import, use skillmeat tag to add collection-level tags
- These persist across refreshes and updates
Artifacts not appearing in Web UI¶
Problem: Imported artifacts visible in CLI but not in Web UI.
Solution:
- Refresh the browser page (hard refresh: Cmd+Shift+R or Ctrl+Shift+R)
- Restart the web server: skillmeat web dev (stop and restart)
- Check browser console for errors (F12 → Console tab)
Best Practices¶
Scan Regularly¶
- After new feature work — Scan to capture new artifacts
- Monthly — Routine audit of what exists locally
- Before cleanup — Identify what can be safely archived
Use Consistent Tags¶
Establish a tagging scheme early and use it consistently:
# Good: Consistent, descriptive
skillmeat tag my-skill --add python,productivity,core
# Less useful: Vague or inconsistent
skillmeat tag my-skill --add thing,stuff
Document Metadata¶
When organizing imported artifacts, include: - Description — What the artifact does - Tags — Category, language, purpose - Owner — Who maintains it - Version — Current version number
This helps future discovery and team collaboration.
Create Snapshots Before Bulk Import¶
If importing many artifacts, create a backup first:
This lets you rollback if something unexpected happens.
Next Steps¶
Now that your artifacts are scanned, imported, and organized:
- Adding Artifacts Guide — Add new artifacts from other sources
- Deploying Artifacts Guide — Deploy to multiple projects
- Bundle & Composite Authoring — Group related artifacts
- Marketplace Guide — Publish or find community artifacts
- Syncing Changes — Keep upstream sources in sync