Skip to content

Walkthroughs available

See the Knowledge Worker Walkthrough for marketplace browsing and deployment, the Product Manager Walkthrough for team publishing, or the Team / Enterprise Admin Walkthrough for governance-aware publishing.

Marketplace — Search, Install, Publish, Lifecycle

The SkillMeat marketplace is a curated repository of verified artifact bundles created by the community. This guide covers the complete marketplace workflow: discovering and installing artifacts as a consumer, publishing as a publisher, and managing your published bundles through their full lifecycle.

Overview

Two roles. Two flows.

Role Purpose Key Tasks
Consumer Find and install verified artifacts Search, install, manage versions, report issues
Publisher Share your artifacts with the community Bundle, publish, maintain versions, deprecate

What You Can Install

The marketplace contains four main artifact types:

  • Skills — Reusable Claude capabilities (text analysis, data processing, code generation)
  • Commands — Automated workflows for common tasks
  • Agents — Specialized subagents for specific domains
  • MCP Servers — Model Context Protocol servers for external integrations

These are distributed as bundles — curated collections of one or more artifacts, signed and verified by SkillMeat.

Trust & Security

All marketplace bundles include:

  • Cryptographic signatures — Verify authenticity and prevent tampering
  • Security scanning — Automated checks for malicious content, hardcoded secrets, and unsafe patterns
  • License verification — Ensure compatibility across all artifacts
  • Publisher ratings — Community feedback and historical performance
  • Provenance tracking — Know where artifacts come from and who published them

Prerequisites

For Consumers (Searchers & Installers)

  • SkillMeat installed — Get started: Quickstart Guide
  • Collection initialized — Run skillmeat init or see Collection Init & Config
  • Access to marketplace — Via Web UI (http://localhost:3000/marketplace) or CLI (skillmeat marketplace-search)

For Publishers (Authors)

In addition to consumer prerequisites:

  • Author credentials — Name and email for publisher attribution
  • Repository source — GitHub repository containing your artifacts (or bundle path)
  • Artifact quality standards — Artifacts must meet SkillMeat validation criteria:
  • Complete documentation
  • Passing security scan (no secrets, safe patterns)
  • Clear license (SPDX identifier)
  • Semantic versioning (major.minor.patch)
  • Understanding of bundle format — Create bundles via CLI or configuration file

Enterprise Context

  • Publisher Approval Workflow — Enterprise deployments may require approval before publishing
  • Role-Based Access — Only designated roles can publish or approve
  • Governance Configuration — See Enterprise Governance

Search & Discover

Basic search:

skillmeat marketplace-search python

Search with filters:

# By type
skillmeat marketplace-search --type skill --query "data processing"

# By tag
skillmeat marketplace-search --tags "productivity,automation"

# By license
skillmeat marketplace-search --license MIT,Apache-2.0

# By author
skillmeat marketplace-search --author "john@example.com"

# Combined filters
skillmeat marketplace-search "document" --type skill --tags "productivity" --limit 20

Browse categories:

# List trending bundles
skillmeat marketplace-search --trending --limit 10

# List recently published
skillmeat marketplace-search --recent --limit 10

# List by category
skillmeat marketplace-search --category productivity

Sorting:

# By popularity (downloads)
skillmeat marketplace-search "python" --sort downloads --order desc

# By recent update
skillmeat marketplace-search "python" --sort updated --order desc

# Alphabetical
skillmeat marketplace-search "python" --sort name --order asc

Web UI Browse & Search

Access the marketplace at: http://localhost:3000/marketplace

Main sections:

  1. Featured — Curated recommended bundles by SkillMeat team
  2. Trending — Most popular this week
  3. New — Recently published (last 30 days)
  4. My Installs — Bundles you've installed (personalized)
  5. Starred — Bundles you've bookmarked

Browsing views:

  • Grid — Visual cards with screenshots and ratings
  • List — Detailed table with metadata
  • Compact — Minimal listings (for high-volume browsing)

Filters in sidebar:

  • Type (skill, command, agent, mcp_server)
  • License (MIT, Apache-2.0, GPL-3.0, etc.)
  • Tags (category-based filtering)
  • Rating (4+ stars, 3+ stars, etc.)
  • Author (search by publisher name)

Search bar:

  • Full-text search across titles, descriptions, and tags
  • Results update in real-time as you type
  • Filter results after searching

Bundle card information:

  • Title and description
  • Author/publisher name with rating
  • Downloads and active installs
  • Tags and license
  • "Install" button (one-click install to collection)
  • "View Details" for full bundle page

Release Information

When browsing a marketplace source on the details page, you can view release information from the upstream GitHub repository:

Latest Release Display: - Release tag — The version identifier (e.g., v2.1.0) - Publication date — When the release was published on GitHub - Release notes — Summary of changes in this release - Assets — Downloadable artifacts (if the release includes them)

Viewing all releases:

Click the Releases tab on a source detail page to see: - Complete chronological list of all releases - Release dates in descending order (most recent first) - Draft and pre-release indicators - Quick filters to show/hide drafts and pre-releases

Release selection for installations:

When installing a marketplace artifact with release tracking enabled, SkillMeat automatically selects the latest stable release based on your binding's pin policy:

  • track_latest_release — Automatically follows the latest non-draft, non-prerelease release
  • snapshot_on_change — Takes a snapshot when changes are detected; relies on manual pin updates
  • track_tags — Tracks available Git tags; selection is manual
  • manual_pin — You manually specify which release to use

For more details on pin policies, see Pin Policy Reference.

Understanding Ratings & Reviews

Star ratings: 1–5 stars based on community feedback

  • 5 stars — Excellent, reliable, well-maintained
  • 4 stars — Good, minor issues or limited scope
  • 3 stars — Okay, useful but with notable limitations
  • 1–2 stars — Issues: poor docs, bugs, or security concerns

Review details: See publication date, author, and full review text on the bundle details page

Your installs: After installing, leave a review to help others decide


Install from Marketplace

Via CLI

Basic install:

skillmeat marketplace-install <bundle-id>

Example:

skillmeat marketplace-install skillmeat-productivity-42

Install specific version:

skillmeat marketplace-install skillmeat-productivity-42 --version 2.1.0

Install with conflict strategy:

# Skip if already installed
skillmeat marketplace-install skillmeat-productivity-42 --strategy skip

# Upgrade to newer version
skillmeat marketplace-install skillmeat-productivity-42 --strategy upgrade

# Overwrite existing
skillmeat marketplace-install skillmeat-productivity-42 --strategy force

View bundle details before installing:

skillmeat marketplace-info skillmeat-productivity-42

# Include reviews and version history
skillmeat marketplace-info skillmeat-productivity-42 --reviews --versions

Installation output:

Fetching bundle from marketplace...
Validating signature...
Scanning security...
Checking licenses...
Installing artifacts:
  ✓ skill-document-processor v1.2.0
  ✓ command-convert-pdf v1.0.0
  ✓ skill-email-automation v1.1.0

Installation complete
  Location: ~/.skillmeat/collection/default/artifacts/
  Version locked in: ~/.skillmeat/collection/default/lock.toml

  1. Search or browse marketplace bundles (see Search & Discover)
  2. Click "Install" on the bundle card or details page
  3. Select collection (if multiple collections exist)
  4. Review permissions warning:

    This bundle includes:
    • 3 skills
    • 2 commands
    • 1 agent
    
    Artifacts will be added to your collection and available
    for deployment to projects immediately after installation.
    
  5. Confirm install — Click "Install Bundle"

  6. Monitor progress:
    • Signature validation (cryptographic verification)
    • Security scan (checks for secrets, malicious patterns)
    • License compatibility check
    • Artifact extraction and placement
  7. Completion confirmation — Shows installed artifacts with versions

Post-Installation Verification

After installing, verify the artifacts are available:

# List installed artifacts
skillmeat list

# Search for a specific installed artifact
skillmeat list --search "document-processor"

# Show version info
skillmeat list --verbose

Output example:

Collection: default
  skill:document-processor v1.2.0 (marketplace/skillmeat-productivity-42)
  command:convert-pdf v1.0.0 (marketplace/skillmeat-productivity-42)
  skill:email-automation v1.1.0 (marketplace/skillmeat-productivity-42)

All installed artifacts are immediately available for deployment. See Deploying Artifacts for next steps.


Publish Your Own Artifact

Prerequisites Checklist

Before publishing, ensure you have:

  • [ ] Artifacts ready — Skills, commands, or agents in your .claude/ project directory
  • [ ] Complete documentation — README with usage examples for each artifact
  • [ ] Passing security scan — No hardcoded secrets (API keys, passwords, tokens)
  • [ ] Valid SPDX license — Choose one: MIT, Apache-2.0, GPL-3.0, BSD-3-Clause, CC0-1.0
  • [ ] Publisher metadata — Your name and email
  • [ ] GitHub repository (optional but recommended) — For source code transparency
  • [ ] Quality checks run — Tests pass, code is formatted, no linting errors
  • [ ] Version tagged — Follow semantic versioning (e.g., 1.0.0)

Step 1: Prepare Metadata

Create metadata file (metadata.json) in your project root:

{
  "title": "Productivity Power Tools",
  "description": "A comprehensive collection of productivity-enhancing Claude skills and commands. Includes document processing, data analysis, automation workflows, and more. Perfect for developers and knowledge workers looking to supercharge their Claude experience.",
  "tags": ["productivity", "automation", "development"],
  "license": "MIT",
  "publisher": {
    "name": "Jane Developer",
    "email": "jane@example.com",
    "homepage": "https://janedeveloper.com"
  },
  "homepage": "https://github.com/jane/productivity-tools",
  "repository": "https://github.com/jane/productivity-tools",
  "documentation": "https://productivity-tools.readthedocs.io",
  "price": 0
}

Required fields:

  • title — 5–100 characters
  • description — 100–5000 characters (detailed, not just a tagline)
  • tags — 1–10 tags from valid list (see Valid Tags)
  • license — SPDX identifier (MIT, Apache-2.0, GPL-3.0, BSD-3-Clause, CC0-1.0, etc.)
  • publisher.name — 1–100 characters
  • publisher.email — Valid email format

Optional fields:

  • publisher.homepage — Publisher's personal/company website
  • homepage — Project landing page or main repository
  • repository — GitHub repository URL
  • documentation — External documentation site
  • price — Price in cents (0 = free, >0 = paid; default: 0)

Step 2: Create a Bundle

Bundle one or more artifacts for release:

# Bundle specific artifacts by name
skillmeat bundle-build productivity-tools \
  --artifact skill:document-processor \
  --artifact command:convert-pdf \
  --artifact skill:email-automation \
  --output productivity-tools.skillmeat-pack

Or from a configuration file:

skillmeat bundle-build --config bundle.toml

See the CLI reference or project documentation for bundling options.

Bundle validation output:

Building bundle: productivity-tools
  Location: productivity-tools.skillmeat-pack
  Size: 2.4 MB
  Artifacts: 3
    - skill:document-processor v1.2.0
    - command:convert-pdf v1.0.0
    - skill:email-automation v1.1.0
  Hash: abc123def456...
  Signature: Signing with publisher key...

Bundle Creation

For detailed bundle authoring, see your project's .claude/ bundle configuration directory or documentation on creating composite artifacts.

Step 3: Validate (Dry Run)

Test publication without submitting:

skillmeat marketplace-publish productivity-tools.skillmeat-pack \
  --title "Productivity Power Tools" \
  --description "A comprehensive collection of productivity-enhancing Claude skills..." \
  --tags "productivity,automation,development" \
  --license "MIT" \
  --publisher-name "Jane Developer" \
  --publisher-email "jane@example.com" \
  --repository "https://github.com/jane/productivity-tools" \
  --dry-run

Expected output (validation-only):

Validating bundle...
  ✓ Bundle integrity (ZIP format, manifest valid)
  ✓ Metadata validation (title, description, tags, license)
  ✓ License compatibility (all artifacts MIT-compatible)
  ✓ Security scan (no secrets, safe patterns)
    - Scanned 47 files
    - No API keys detected
    - No hardcoded passwords found
  ✓ File type validation (safe extensions)
  ✓ Size check (2.4 MB < 100 MB limit)

All validation passed! Ready to publish.

If validation fails, fix the issues and re-run the dry run.

Step 4: Publish

Once dry run passes, publish to the marketplace:

skillmeat marketplace-publish productivity-tools.skillmeat-pack \
  --title "Productivity Power Tools" \
  --description "A comprehensive collection of productivity-enhancing Claude skills and commands..." \
  --tags "productivity,automation,development" \
  --license "MIT" \
  --publisher-name "Jane Developer" \
  --publisher-email "jane@example.com" \
  --repository "https://github.com/jane/productivity-tools"

Publishing output:

Submitting to marketplace...

Submission received
  ID: skillmeat-productivity-42
  Status: pending
  Created: 2026-04-20 14:23:45 UTC
  Review time: Typically 1-3 business days

Track status:
  skillmeat marketplace-status skillmeat-productivity-42

Contact:
  marketplace@skillmeat.com
  1. Navigate to ProfileMy Published Bundles
  2. Click "Publish New Bundle"
  3. Upload bundle file (.skillmeat-pack) or use Publish from Collection
  4. Fill metadata form:
    • Title
    • Description (100+ chars)
    • Tags (select from list)
    • License (dropdown)
    • Publisher info (auto-filled from profile)
    • Optional: Homepage, Repository, Documentation links
  5. Security review: Bundle is scanned automatically
  6. Submit for approval
  7. Confirmation: Shows submission ID and expected review time

Step 5: Track Status & Governance

Publishing is irreversible

Once your bundle is approved and published, it's available to all users. Plan carefully for versioning and updates.

Track submission status:

skillmeat marketplace-status skillmeat-productivity-42

Output:

Submission: skillmeat-productivity-42
Status: in_review (under active review)
Submitted: 2026-04-20 14:23:45 UTC
Reviewer: Sarah M. (SkillMeat Team)
Progress:
  ✓ Metadata validation
  ✓ Security scan
  ✓ License check
  ⏳ Functional testing (in progress)
Expected completion: 2026-04-21 16:00:00 UTC

Next steps:
  - Reviewer may request clarifications or changes
  - You'll receive email notifications on status changes

Possible approval workflows:

Status Meaning What to Do
pending Submitted, awaiting review queue Wait (1–2 days typical)
in_review Under active review Wait for email or status updates
approved Published to marketplace! Available immediately to all users
rejected Declined, review feedback provided Fix issues and resubmit
revision_requested Needs changes before approval Make changes and update submission

Enterprise approval workflow:

In enterprise deployments, additional roles may be required:

  • Publisher submits → bundle enters pending state
  • Publisher approval role reviews → may request revision or approve
  • Compliance/governance checks (if configured) → verify security and license policies
  • Approved → published to marketplace

See Enterprise Governance for role definitions and approval policies.


Publisher Lifecycle

Version Management

SkillMeat uses semantic versioning: MAJOR.MINOR.PATCH

Release types:

Type When to use Example
Major Breaking changes, incompatible API changes 1.0.02.0.0
Minor New features, backward compatible 1.2.01.3.0
Patch Bug fixes, maintenance 1.2.11.2.2
Pre-release Beta/alpha versions 2.0.0-beta.1

Breaking change flag:

When publishing a major version with breaking changes, mark it explicitly:

skillmeat marketplace-publish v2.0.0.skillmeat-pack \
  --title "My Tool v2.0" \
  --breaking-change \
  --description "Version 2.0: Major rewrite with new architecture. See MIGRATION.md for upgrade path."
  # ... other args

This helps users understand the impact and plan upgrades.

Changelog Requirements

For each release, include a changelog entry describing changes:

Format: Keep entries in CHANGELOG.md at your project root following Changelog Conventions:

## [2.0.0] - 2026-04-20

### Added
- New feature X (breaking change)
- Improved performance by 50%

### Changed
- Renamed function Y to Z (breaking)
- Updated dependencies

### Fixed
- Bug: memory leak in processor
- Issue: race condition in sync

### Deprecated
- Function `old_func()` (use `new_func()` instead)
- Parameter `old_param` (removed in v2.1.0)

### Removed
- Support for Python 3.8 (now requires 3.9+)

Include changelog entry in bundle description:

skillmeat marketplace-publish v2.0.0.skillmeat-pack \
  --title "My Tool v2.0" \
  --description "Version 2.0: Major rewrite

## Changelog
- New feature X
- Performance improvements
- Breaking change: Function Y renamed to Z

See CHANGELOG.md for full details." \
  # ... other args

Version Channels

Control which versions users see:

Channel types:

  • latest — Newest version (default for installs)
  • stable — Recommended for production (typically LTS releases)
  • beta — Pre-release versions (opt-in)
  • legacy — Older major versions (for users who can't upgrade yet)

Setting channels via bundle metadata:

skillmeat marketplace-publish v2.0.0.skillmeat-pack \
  --version-channel stable \
  # ... other args

Or update existing published versions via Web UI:

  1. Navigate to My Published BundlesYour Bundle
  2. Click Versions tab
  3. Select version → Set as → Choose channel (latest, stable, beta, legacy)

User perspective:

  • skillmeat marketplace-install <bundle> → installs latest version
  • skillmeat marketplace-install <bundle> --version 1.5.0 → installs specific version
  • Web UI shows all available versions and indicates the current latest/stable channels

Deprecation & Sunset

As a publisher, you control when older versions reach end-of-life.

Deprecate a version:

skillmeat marketplace-deprecate <bundle-id> --version 1.0.0 \
  --message "Version 1.0.0 is deprecated. Please upgrade to v2.0 for bug fixes and performance improvements." \
  --sunset-date 2026-07-20

Effects:

  • Marked deprecated in marketplace listings
  • Installation warning: "This version is deprecated and will be removed on 2026-07-20"
  • Users are encouraged to upgrade
  • On sunset date, version is removed from available installs (old installs still work)

Deprecation timeline best practice:

  1. Release new major version with migration guide
  2. Wait 60 days for users to discover and upgrade
  3. Mark deprecated with sunset date 30 days out (90 days total from release)
  4. Enforce sunset — version removed from available downloads
  5. Users with old installs can still use them; they just can't reinstall from marketplace

Usage Analytics & Insights

Monitor your published bundles:

skillmeat marketplace-analytics <bundle-id>

Output:

Bundle: My Productivity Tools (skillmeat-42)
Published: 2025-12-01

Downloads: 1,247
Active installs: 892
Rating: 4.3/5 (256 reviews)

Version breakdown:
  v2.0.0 (latest): 612 active
  v1.5.0 (stable): 178 active
  v1.0.0 (deprecated): 102 active

Install trend (last 30 days):
  Week 1: 145 downloads
  Week 2: 128 downloads
  Week 3: 92 downloads
  Week 4: 53 downloads

Top feedback:
  ✓ Easy to use (95% positive)
  ✓ Good documentation (87% positive)
  ⚠ Performance issues in v2.0 (12 reports)

Via Web UI:

Navigate to My Published BundlesYour BundleAnalytics tab to see:

  • Download and install trends (charts)
  • Active installs by version
  • User reviews and ratings
  • Common support issues
  • Version upgrade patterns

Governance & Permissions

Publisher Roles

Role Permissions When to assign
Author Create bundles, submit for review, view own analytics All publishers
Publisher Also: update published bundles, set version channels, manage deprecation Trusted authors with multiple published bundles
Marketplace Admin Approve submissions, manage community standards, suspend/remove listings SkillMeat team + enterprise admins

Review & Approval Process

SkillMeat Hosted (Community):

  1. You submit bundle → enters pending state
  2. SkillMeat team reviews (1–3 business days):
  3. Validate metadata and security scan
  4. Check for malicious patterns or license conflicts
  5. Verify code quality and documentation
  6. Approved → published immediately
  7. Rejected → receive feedback and can resubmit after fixes

Enterprise Deployment:

  1. You submit bundle → awaits approval role
  2. Approval role (e.g., "marketplace-reviewer") reviews
  3. If configured: compliance/governance checks run
  4. Approved → published to enterprise marketplace
  5. Rejected → receive feedback

See Enterprise Governance for role setup and approval policies.

Content Guidelines

To ensure quality and trust:

  • No spam or duplicates — Don't republish the same artifact multiple times
  • Accurate descriptions — Clearly describe what your bundle does
  • Safe code — No malware, obfuscation, or intentional security risks
  • Respect licenses — All artifacts must be properly licensed and compatible
  • No impersonation — Don't claim authorship of others' work
  • Responsive support — Reply to user questions and issues

Violations may result in bundle removal and marketplace access suspension.


Troubleshooting

Installation Issues

"Bundle not found"

Error: Bundle 'skillmeat-unknown-99' not found in marketplace

Solution: Verify the bundle ID using search:

skillmeat marketplace-search "keyword"

Copy the exact bundle ID from results and retry.

"Permission denied: insufficient permissions"

In enterprise deployments, you may lack permission to install from certain publishers.

Solution: Contact your enterprise admin to grant marketplace install permissions.

"Signature validation failed"

Bundle may be corrupted or tampered with.

Solution:

  1. Retry the install (may be temporary network issue)
  2. Clear local cache: skillmeat cache clear
  3. Contact support if issue persists

"License conflict: GPL artifact in MIT bundle"

Solution: This is a warning (not an error). Understand the implications:

  • A GPL artifact requires your entire project to be GPL-compatible
  • If you need MIT-only, choose a different bundle or contact the publisher

Publishing Issues

"Security violation: AWS key detected"

Error: Potential AWS key found in skill/config.py line 42

Solution:

  1. Remove the hardcoded key
  2. Use environment variable instead: os.environ.get("AWS_KEY")
  3. Verify with git-secrets or similar tool
  4. Rerun dry-run validation

"Description too short"

Error: Description must be at least 100 characters (you provided 45)

Solution: Expand your description with details about features and use cases.

"Invalid license identifier"

Error: 'My Custom License' is not a valid SPDX identifier

Solution: Use a standard SPDX license. Common options:

  • MIT — Permissive, most flexible
  • Apache-2.0 — Permissive with patent grant
  • GPL-3.0-only — Copyleft, requires source disclosure
  • BSD-3-Clause — Permissive with attribution

Full list: SPDX License List

"Submission rejected: insufficient documentation"

Solution: Add clear documentation to each artifact:

  • README with usage examples
  • Comments for complex logic
  • Type hints (Python) or JSDoc (JavaScript)
  • Link to external docs if available

Resubmit after improvements.

Version Management Issues

"Version already exists"

Error: Version 1.0.0 already published (you cannot overwrite)

Solution: Versions are immutable once published (prevents breaking user installs). Create a new patch release:

# Instead of republishing 1.0.0, publish 1.0.1
skillmeat marketplace-publish v1.0.1.skillmeat-pack ...

"Can't deprecate: version has active installs"

Solution: You can deprecate even with active installs. This is the point — warn users to upgrade:

skillmeat marketplace-deprecate <bundle> --version 1.0.0 \
  --message "Version 1.0.0 has a critical bug. Upgrade to v1.0.1 or v2.0." \
  --sunset-date 2026-05-20

Active installs continue to work; new installs are blocked.

Governance & Approval

"Submission awaiting approval, takes too long"

SkillMeat typical review time is 1–3 business days. Enterprise may have custom workflows.

Solution: Monitor status:

skillmeat marketplace-status <submission-id>

If you don't see progress after 3 days, contact: - SkillMeat Hosted: marketplace@skillmeat.com - Enterprise: Your marketplace admin

"Revision requested: need to make changes"

A reviewer found issues and wants you to improve before approval.

Solution:

  1. Check submission details: skillmeat marketplace-status <id>
  2. Read reviewer feedback
  3. Make requested changes
  4. Create new bundle version
  5. Resubmit: skillmeat marketplace-publish updated.skillmeat-pack ...

Getting Help

For consumers:

  • Marketplace issue: Open issue in the bundle's GitHub repository
  • Installation problem: Contact SkillMeat support

For publishers:

  • Publication question: marketplace@skillmeat.com
  • Security concern: security@skillmeat.com (preferred for sensitive issues)
  • Feature request: Community forum or GitHub issues

Valid Tags

Use one or more of these tags when publishing:

productivity       automation        development       testing
data-analysis      documentation     ai-ml             web-dev
backend            frontend          database          security
devops             cloud             api               cli
education         research          creative          business

Next Steps

As a consumer:

As a publisher:


Summary

Task CLI Command Web UI Path
Search skillmeat marketplace-search <query> /marketplace + search bar
Install skillmeat marketplace-install <bundle-id> /marketplace → "Install"
Publish skillmeat marketplace-publish <file> Profile → My Published Bundles
Check status skillmeat marketplace-status <id> Profile → My Published Bundles
View analytics skillmeat marketplace-analytics <id> Profile → My Published Bundles → Analytics
Deprecate skillmeat marketplace-deprecate <id> My Published Bundles → Versions