Skills Dojo API

15 live skills across 11 blockchains. Execute any skill with a single POST request.

Quick Start

# List all skills

curl /api/skills

# Execute a skill

curl -X POST /api/execute \
-H "Content-Type: application/json" \
-d '{"skillId":"gas-tracker-agent","params":{}}'

Endpoints

GET

/api/skills

List all skills. Query params: ?category=, ?chain=, ?q=, ?format=compact

GET

/api/skills/{id}

Full skill detail with parameter schema and examples

POST

/api/execute

Execute a skill. Body: { "skillId": "...", "params": { ... } }

GET

/api/agent-spec

LLM-consumable skill catalog (markdown). Use ?format=json for structured JSON

Response Format

All execution responses use a consistent envelope:

{
  "ok": true,
  "skillId": "gas-tracker-agent",
  "executionTimeMs": 1230,
  "result": {
    "status": "success",
    "message": "...",
    "data": { ... },
    "explorerUrl": "https://...",
    "timestamp": "2026-02-12T..."
  }
}

On error: {"ok": false, "skillId": "...", "error": "message"} with HTTP 4xx/5xx.

For LLM Agents

Fetch the full skill catalog as a single document your LLM can use as context:

# Markdown (for system prompts)

curl /api/agent-spec

# JSON (for tool definitions)

curl /api/agent-spec?format=json

The spec includes all 15 skill definitions with parameters, defaults, and example payloads. Feed the markdown to your LLM as a system prompt to enable skill execution.

Rate Limits

10 requests/hour per IP for /api/execute. Skill listing and spec endpoints are cached and not rate-limited.

Exceeding the limit returns HTTP 429: {"ok": false, "error": "Rate limit exceeded..."}

Skill Catalog

Click any skill to see its parameters and a ready-to-use cURL example.

Skill IDCategoryChainsParams