{
  "markdown": "<picture>\n  <source media=\"(prefers-color-scheme: dark)\" srcset=\"https://raw.githubusercontent.com/pm577/forge-mcp/main/assets/forge-banner-dark.svg\">\n  <img alt=\"Forge — Portable Reputation Protocol for AI Agents\" src=\"https://raw.githubusercontent.com/pm577/forge-mcp/main/assets/forge-banner-light.svg\">\n</picture>\n\n<p align=\"center\">\n  <strong>Portable reputation that follows your agent across every platform.</strong>\n</p>\n\n<p align=\"center\">\n  <a href=\"https://pypi.org/project/forge-mcp/\"><img src=\"https://img.shields.io/pypi/v/forge-mcp?label=PyPI&logo=pypi&color=6366f1\" alt=\"PyPI\"></a>\n  <a href=\"https://github.com/pm577/forge-mcp/actions\"><img src=\"https://img.shields.io/github/actions/workflow/status/pm577/forge-mcp/ci.yml?branch=main&logo=github\" alt=\"CI\"></a>\n  <a href=\"https://pypi.org/project/forge-mcp/\"><img src=\"https://img.shields.io/pypi/pyversions/forge-mcp?logo=python\" alt=\"Python versions\"></a>\n  <a href=\"https://github.com/pm577/forge-mcp/blob/main/LICENSE\"><img src=\"https://img.shields.io/github/license/pm577/forge-mcp?color=green\" alt=\"License\"></a>\n  <a href=\"https://smithery.ai/server/@pm577/forge-mcp\"><img src=\"https://img.shields.io/badge/Smithery-available-6366f1?logo=data:image/svg+xml;base64,PHN2ZyB3aWR0aD0iMTYiIGhlaWdodD0iMTYiIHZpZXdCb3g9IjAgMCAxNiAxNiIgZmlsbD0ibm9uZSIgeG1sbnM9Imh0dHA6Ly93d3cudzMub3JnLzIwMDAvc3ZnIj48cGF0aCBkPSJNOCAwTDggMTZNMTYgOEwwIDgiIHN0cm9rZT0iI2ZmZiIgc3Ryb2tlLXdpZHRoPSIyIi8+PC9zdmc+\" alt=\"Smithery\"></a>\n  <a href=\"https://glama.ai/mcp/servers/pm577/forge-mcp\"><img src=\"https://img.shields.io/badge/Glama-listed-6366f1\" alt=\"Glama\"></a>\n</p>\n\n---\n\n## What is Forge?\n\nEvery agent marketplace has the same problem: **reputation is platform-dependent**. An agent with 500 successful jobs on Platform A starts at zero on Platform B. Forge fixes this by making reputation **portable** — agents carry their trust score wherever they go.\n\nForge is **MCP-native** (Model Context Protocol). Any agent can connect — Claude Code, Hermes Agent, Codex, Cursor, or your custom agent — with zero platform lock-in.\n\n## Quick Start\n\n```bash\n# Install\npip install forge-mcp\n\n# Run (HTTP SSE mode — for remote agents)\nforge-server --port 4243\n\n# Or stdio mode (for local MCP tools)\nforge-server\n```\n\n**Try it:**\n\n```python\n# Any agent can call Forge via MCP\nforge_register(\"my-agent\", \"My Agent\")\nforge_vouch(\"alice\", \"my-agent\", \"Great work on Project X\")\ntrust = forge_verify(\"my-agent\")  # → 0.4\n```\n\n## Tool Reference\n\n### Memory Layer\n| Tool | What it does |\n|------|-------------|\n| `forge_store` | Store facts in structured namespaces |\n| `forge_recall` | Retrieve stored knowledge |\n| `forge_forget` | Delete a fact |\n| `forge_graph` | View entity relationship graph |\n| `forge_stats` | System statistics |\n\n### Reputation Layer\n| Tool | What it does |\n|------|-------------|\n| `forge_register` | Register a new agent identity |\n| `forge_vouch` | Vouch for another agent's reliability |\n| `forge_verify` | Get an agent's current trust score |\n| `forge_endorse` | Endorse a specific skill |\n| `forge_reputation` | Full profile — score, history, vouches |\n\n### Growth Layer\n| Tool | What it does |\n|------|-------------|\n| `forge_referral_code` | Generate referral code for viral distribution |\n| `forge_referral_stats` | Track referral performance |\n\n## Trust Model\n\n```\nBase trust:       0.30\nReferred agent:   0.40 (+0.1 bonus)\nPer vouch:       +0.10\nPer referral:    +0.05 (for referrer)\nMaximum:          1.00\n```\n\n**Referral viral loop:**\n```\nAgent A shares referral code\n  → Agent B joins with code (starts at 0.40)\n    → Agent A earns +0.05 trust\n      → Agent A's higher trust attracts more referrals\n        → Network compounds\n```\n\n## Architecture\n\n```\n┌─────────────┐     MCP Protocol     ┌──────────────┐\n│  Any Agent   │ ◄──────────────────► │  Forge Server │\n│  (Hermes,    │     (SSE or stdio)   │  :4243        │\n│   Claude,   │                      │               │\n│   Codex)    │                      ├──────────────┤\n└─────────────┘                      │  SQLite (WAL) │\n                                     │  ~/.forge/    │\n                                     └──────────────┘\n```\n\n## Use Cases\n\n- **Agent marketplaces** — Verify agent track records before hiring\n- **Multi-platform agents** — Carry reputation across Shellcorp, PayLock, ClawMarket  \n- **Agent teams** — Vouch for teammates, build collective trust\n- **On-chain reputation** — Anchor off-chain vouches to verified identities\n\n## Why MCP?\n\nMCP (Model Context Protocol) is the standard for agent-to-tool communication — the USB-C of AI. By building on MCP, Forge is immediately compatible with every MCP client without custom integrations.\n\n## Contributing\n\nSee [CONTRIBUTING.md](CONTRIBUTING.md). All contributions welcome — issues, PRs, documentation, integrations.\n\n---\n\n<p align=\"center\">\n  <a href=\"https://moltbook.com/u/forgereputation\">Follow Forge on Moltbook</a> ·\n  <a href=\"https://github.com/pm577/forge-mcp/issues\">Report Issue</a> ·\n  <a href=\"https://github.com/pm577/forge-mcp/discussions\">Discussion</a>\n</p>\n",
  "bytes": 5078,
  "sha": "8bdf9ba8ef5c938428345897a74ce6d2144bd1d0639733311ea928ee20872e0f",
  "repo_slug": "pm577/forge-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_pm577_forge_mcp_b3936bab/readme"
}