MetEngine Data Agent Skill

LLM skill for MetEngine's x402 pay-per-request analytics API across:

• Polymarket
• Hyperliquid
• Meteora

The repo is structured for progressive disclosure so agents load minimal context first, then fetch only the docs needed for the active platform.

Repo Layout

• SKILL.md: compact orchestrator instructions and loading rules
• agents/openai.yaml: listing metadata for skill directories/UIs
• references/docs-index.json: machine-readable routing map
• references/docs-index.md: human-readable routing guide
• references/core-runtime.md: minimal runtime payment flow
• references/core-extended.md: advanced/troubleshooting details
• references/polymarket-endpoints.md: Polymarket endpoint reference
• references/hyperliquid-endpoints.md: Hyperliquid endpoint reference
• references/meteora-endpoints.md: Meteora endpoint reference

Design Goals

• Keep always-loaded skill context small
• Route queries to one platform doc by default
• Support remote raw-doc loading from GitHub when skill is installed elsewhere
• Preserve strict output rule: never truncate addresses/IDs

Quick Local Check

1. Open SKILL.md.
2. Load references/docs-index.json.
3. Load references/core-runtime.md.
4. Load exactly one platform file based on user intent.

Remote Doc Endpoints

• https://raw.githubusercontent.com/MetEngine/skill/main/references/docs-index.json
• https://raw.githubusercontent.com/MetEngine/skill/main/references/core-runtime.md
• https://raw.githubusercontent.com/MetEngine/skill/main/references/core-extended.md
• https://raw.githubusercontent.com/MetEngine/skill/main/references/polymarket-endpoints.md
• https://raw.githubusercontent.com/MetEngine/skill/main/references/hyperliquid-endpoints.md
• https://raw.githubusercontent.com/MetEngine/skill/main/references/meteora-endpoints.md