Skip to content
Back to search

anthrope-agent

Onchain intelligence for traders and AI agents: live terminal, AI analyst, data API and MCP server.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/anthropeagent--anthrope-agent/badge.svg)](/listings/anthropeagent--anthrope-agent)

HTML

<a href="/listings/anthropeagent--anthrope-agent"><img src="/api/listings/anthropeagent--anthrope-agent/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 121 lines
  1. <p align="center">
  2. <img src=".github/assets/banner.jpg" alt="Anthrope Agent: read the chain before the crowd" width="100%" />
  3. </p>
  4. <p align="center">
  5. <img src=".github/assets/logo.png" alt="Anthrope Agent logo" width="72" />
  6. </p>
  7. <h1 align="center">Anthrope Agent</h1>
  8. <p align="center">
  9. <b>Onchain intelligence for traders and AI agents.</b><br />
  10. A live market terminal, an AI analyst, a data API and an MCP server, on one domain.
  11. </p>
  12. <p align="center">
  13. <a href="https://www.npmjs.com/package/anthrope-agent-mcp"><img alt="npm" src="https://img.shields.io/npm/v/anthrope-agent-mcp?color=e8d9bc&labelColor=17120e&label=anthrope-agent-mcp" /></a>
  14. <a href="https://x.com/anthropeagent"><img alt="X" src="https://img.shields.io/badge/X-@anthropeagent-e8d9bc?labelColor=17120e&logo=x&logoColor=f3ecdf" /></a>
  15. <img alt="Node" src="https://img.shields.io/badge/node-%3E%3D20-e8d9bc?labelColor=17120e" />
  16. <img alt="TypeScript" src="https://img.shields.io/badge/TypeScript-strict-e8d9bc?labelColor=17120e&logo=typescript&logoColor=f3ecdf" />
  17. </p>
  18. ---
  19. ## What it is
  20. Anthrope Agent reads public blockchains and DEX markets and turns them into something a person or an AI agent can act on: prices, liquidity, holders, security scans, wallet profiles, top traders, pump.fun trenches and Hyperliquid perps.
  21. It is **read-only and non-custodial**. It never holds your keys or funds; every swap is signed in your own wallet.
  22. | Product | What you get |
  23. | --- | --- |
  24. | **Terminal** | Discover (trending, new, top market cap, RWA), token pages with live charts, trades, holders and security, wallet profiles, top traders, trenches, perps and alerts. Updates stream over WebSocket. |
  25. | **AI Analyst** | Ask in plain words ("is this token safe?"). Claude reads the live data and explains what it sees. Runs in your browser with your own Anthropic key, which never touches our servers. |
  26. | **MCP server** | [`anthrope-agent-mcp`](apps/mcp): 13 read-only tools that give Claude Desktop, Claude Code, Cursor and any MCP client the same live data. |
  27. | **Data API** | REST `/api/v1` and WebSocket `/ws` with API keys and per-plan rate limits. |
  28. | **Research** | Reports and blog posts, managed from a built-in CMS at `/admin`. |
  29. | **Peak** | Institutional dashboard modules built on the same data. |
  30. ## MCP server in 30 seconds
  31. ```sh
  32. claude mcp add anthrope-agent \
  33. -e ANTHROPE_API_KEY=ak_... \
  34. -e ANTHROPE_API_URL=https://YOUR-DOMAIN/api/v1 \
  35. -- npx -y anthrope-agent-mcp
  36. ```
  37. Then ask Claude: *"Check BONK on Solana: security, holder concentration and what top traders are doing."*
  38. Full tool list and Claude Desktop / Cursor config: [apps/mcp/README.md](apps/mcp/README.md).
  39. ## Architecture
  40. ```text
  41. browser ──► Next.js (apps/web) ──► /api/v1, /ws ──► Fastify (apps/api)
  42. │
  43. GeckoTerminal, DEX Screener, GoPlus, workers (apps/workers) ─┤──► Postgres · Redis · ClickHouse
  44. Hyperliquid, pump.fun, Jupiter, Kyber ──► provider bridge ────────┘
  45. ▲
  46. Claude / Cursor ──► MCP server (apps/mcp) ──── x-api-key ─────────┘
  47. ```
  48. ```text
  49. apps/
  50. web/ Next.js 15 (App Router, React 19, Tailwind v4): hub, terminal, analyst, research, admin
  51. api/ Fastify REST /api/v1 and WebSocket /ws, wallet sign-in, API keys, CMS, swap builder
  52. workers/ Background jobs: provider bridge, scans, trenches, perps, alerts
  53. mcp/ MCP server (published to npm as anthrope-agent-mcp)
  54. packages/
  55. types/ Shared types
  56. db/ Postgres / ClickHouse / Redis clients and SQL migrations
  57. bridge/ Provider adapters, normalized into one token model
  58. devdb/ Postgres (PGlite) + in-memory Redis for local development without Docker
  59. infra/ docker-compose and Caddyfile for a single-domain production setup
  60. docs/ Architecture, API contract (OpenAPI + WebSocket), development guide
  61. ```
  62. ## Run it locally
  63. Requires **Node.js 20+**. Docker is optional.
  64. ```sh
  65. npm install
  66. cp .env.example .env
  67. # with Docker
  68. npm run db:up
  69. # or without Docker (Windows without WSL)
  70. npm run db:local # keep this running in its own terminal
  71. npm run db:migrate
  72. npm run dev # web :3000, api :4000, workers
  73. ```
  74. Open <http://localhost:3000>. The first live data arrives about a minute after the workers start. To run the UI with sample data and no database, set `NEXT_PUBLIC_MARKET_DATA=mock` in `.env`.
  75. | Command | What it does |
  76. | --- | --- |
  77. | `npm run typecheck` | Typecheck every workspace |
  78. | `npm run lint` | Lint the web app |
  79. | `npm test` | Run all tests |
  80. | `npm run build` | Production build of web and api |
  81. More detail (data sources, limits per provider, admin setup) is in [docs/DEVELOPMENT.md](docs/DEVELOPMENT.md) and [docs/ARCHITECTURE.md](docs/ARCHITECTURE.md).
  82. ## Security
  83. - Wallet sign-in uses a signed message (SIWS / SIWE). No transaction, no fee, no private key.
  84. - Swaps are built and simulated server-side, then signed in your wallet. Platform fee is 0%.
  85. - API keys are stored only as SHA-256 hashes.
  86. - Secrets live in `.env`, which is never committed. See `.env.example` for every variable.
  87. Found a vulnerability? Please report it privately via a [GitHub security advisory](../../security/advisories/new) instead of a public issue.
  88. ## Disclaimer
  89. Market data comes from third parties and can be delayed or wrong. Security scores are automated checks, not audits. Nothing here is investment advice.
  90. Anthrope Agent is an independent project and is not affiliated with or endorsed by Anthropic. Claude is a trademark of Anthropic. Hero artwork: Raphael, *The School of Athens* (public domain).
  91. <p align="center"><sub>© 2026 Anthrope Agent · <a href="https://x.com/anthropeagent">@anthropeagent</a></sub></p>

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.