The Skill Lifecycle: From Discovery to Production

Browse by Category

The catalog is organized by category — Data, AI/ML, Web Scraping, Communication, Security, and more. Each skill card shows the name, price, success rate, and trust score at a glance.

Search by Use Case

Use the search bar to find skills by what they do. For example, search "scrape website" or "translate text" to find relevant skills.

CLI Discovery

Use the OpenClaw CLI to search from your terminal:

Agent Auto-Discovery

AI agents can discover skills programmatically via the API. Each skill page includes structured data (JSON-LD) and agent-intent metadata so LLMs can match skills to tasks during planning.

Key Metrics

Four numbers tell you the most:

• Trust Score — Overall quality rating (0–100), based on reliability, response time, and usage
• Success Rate — Percentage of calls that return a 2xx response
• Total Calls — How many times this skill has been used (social proof)
• Avg Response Time — Typical latency in milliseconds

Live Output Preview

Each skill shows a real or simulated API response so you can see exactly what data you'll get back before spending anything. Look for the "200 OK" preview panel on the skill page.

Input/Output Schema

The Technical Details section shows the exact JSON schema for inputs and outputs. This tells you what parameters to send and what structure to expect back.

Pricing Transparency

Every skill shows its per-call price upfront. Some skills are free, some offer a free tier with paid overage, and some are pay-per-call from the first request. Failed calls (4xx/5xx/timeout) are never charged.

SKILL.md Documentation

Skills that follow the open-source SKILL.md standard include rich documentation: when to use, prerequisites, API call examples, error handling, and pricing — all rendered directly on the skill page.

Automated Security Scan (OSV)

Every skill with a public repository is automatically scanned for known vulnerabilities using the OSV database. The scan checks all dependencies for CVEs and known exploits. Results are shown as:

• No Vulnerabilities — All dependencies are clean
• Vulnerabilities Found — Known issues detected, with links to details
• Source Unavailable — No public repo to scan (closed-source skill)

Open Source Verification

Skills with a linked GitHub repository show a "Source" link so you can inspect the code yourself. The license type (MIT, Apache 2.0, etc.) is displayed as a badge.

Call History Chart

The Trust section includes a real-time bar chart of recent API calls showing response times and success/failure status. This gives you a live view of the skill's operational health.

Risk Reversal

Claw0x enforces a "no charge on failure" policy. If a skill returns a 4xx error, 5xx error, or times out, you are not billed. Only successful 2xx responses are charged. This removes the financial risk of trying new skills.

Live Sandbox Testing

Every skill page includes a Live Sandbox where you can test real API calls with your own API key before deploying. Enter your input, click "Test", and see the actual response.

Error Handling

Skills return standard HTTP status codes. Handle these in your code: 401 (bad API key), 402 (insufficient balance), 429 (rate limit), 502 (skill backend error). See Error Codes for details.

Monitor Usage

Track your API calls, spending, and per-skill usage from the Dashboard → Calls page. Each call shows the skill, cost, response time, and status.

Balance Management

Keep your USDC balance topped up to avoid 402 errors. Set up a routine to check your balance via the dashboard or API. See the Funding Guide for deposit methods.

Skill Composition

Skills can be chained together. For example: scrape a webpage → analyze sentiment → translate the summary. The "Often Used With" section on each skill page suggests complementary skills for common workflows.