Skip to content

SKILL.md

Latest commit

 

History

History
49 lines (33 loc) · 2.2 KB

SKILL.md

File metadata and controls

49 lines (33 loc) · 2.2 KB
Raw
name tempo-docs
description Answer Tempo blockchain questions using official documentation. Use when asked about Tempo protocol, TIP-20 tokens, fees, transactions, stablecoin DEX, or any Tempo-related questions.

Tempo Docs

Skill for navigating Tempo documentation.

Current network facts

  • Tempo Mainnet has been live since March 18, 2026. Its chain ID is 4217.
  • Tempo Wallet uses mainnet for balances, funding, transfers, and MPP settlement.
  • pathUSD is a live mainnet TIP-20 stablecoin issued by Bridge at 0x20c0000000000000000000000000000000000000.
  • Moderato is Tempo Testnet, with chain ID 42431. Faucet-issued pathUSD on testnet is a test token, not the production asset.

MCP server

Use the hosted MCP server at https://mcp.tempo.xyz first. It exposes these tools:

Tool Description
search Search Tempo and related documentation
find_pages Find matching page URLs from a source index
read_page Read one cleaned documentation page
code Run multi-step documentation lookups

If a user reports stale, missing, or confusing Tempo docs while using MCP context, send sanitized feedback to https://tempo.xyz/developers/api/feedback with source: "mcp", message, and any relevant toolName or relatedResource.

Direct fallbacks

If MCP is unavailable, fetch context directly:

  • llms.txt – Concise index of all pages: https://tempo.xyz/developers/llms.txt
  • Markdown pages – Append .md to any page URL (e.g. https://tempo.xyz/developers/quickstart/integrate-tempo.md)

Use read_web_page to fetch these when you need broad context or a quick answer.

Documentation sources

The MCP server searches Tempo, Viem, Wagmi, Vocs, MPP, Tempo Accounts, TIPs, and Regen documentation. Use the repository's own code tools when you need exact source files.

Workflow

  1. Search docs: Use search to find relevant context
  2. Find pages: Use find_pages when you need canonical page URLs
  3. Read pages: Use read_page with the returned source and path or URL
  4. Combine lookups: Use code for multi-step documentation queries
  5. Fallback: If MCP is unavailable, read llms.txt or fetch a specific page as Markdown