Collections API & Curation Methodology

Browse curated skill collections programmatically, understand how skills are evaluated, and deploy expert-built agent templates.

Getting Started

ClawZenith provides curated skill collections and expert templates via a simple API. Browse collections, install tested skill sets, and deploy production-ready agents.

1. Install the CLI

npm install -g @clawzenith/cli

2. Browse Collections

# List all curated collections

clawzenith collections list

# View a specific collection

clawzenith collections show productivity-suite

3. Install a Collection

# Install all skills in a collection to your OpenClaw instance

clawzenith install productivity-suite

# Deploy an expert template

clawzenith templates deploy sales-agent

Collections API

Programmatic access to browse, search, and install curated collections.

List Collections

GET /api/v1/collections

Authorization: Bearer <api-key>

Response:

[{ "slug": "productivity-suite", "name": "Productivity Suite",

"skills_count": 12, "rating": 4.9, "category": "productivity" },

{ "slug": "dev-toolkit", "name": "Developer Toolkit",

"skills_count": 8, "rating": 4.8, "category": "development" }]

Get Collection Details

GET /api/v1/collections/:slug

Response:

{

"slug": "productivity-suite",

"skills": [{ "name": "calendar-sync", "version": "2.1.0", "tested": true }, ...],

"compatibility": "openclaw >= 3.0",

"last_verified": "2026-03-20"

}

Install Collection

POST /api/v1/collections/:slug/install

{

"target": "https://my-openclaw.example.com",

"overwrite_existing": false

}

Curation Methodology

Every skill in a ClawZenith collection passes a rigorous evaluation process before inclusion.

1. Security audit

Static analysis, dependency review, and permission scope verification. No network access beyond declared APIs.

2. Compatibility testing

Tested against OpenClaw 3.x+ and verified to work with every other skill in the same collection.

3. Performance benchmarks

Response time, memory usage, and token consumption measured. Skills that degrade agent performance are excluded.

4. Expert review

OpenClaw power users evaluate real-world utility, prompt quality, and edge case handling.

Collections are re-verified monthly. Skills that break compatibility or fail updated security checks are removed and subscribers are notified.

Ready for the premium OpenClaw experience?

Try curated collections and expert templates today.