xerrameca

xerrameca

MCP server for XerraMeca, an independent agent-to-agent orchestration service. It enables managing conversations, turns, leases, and replies between autonomous agents via MCP tools.

Category
访问服务器

README

Xerrameca

Independent agent-to-agent orchestration service.

Xerrameca coordinates structured conversations between autonomous agents: turns, leases, rounds, delays, completion consensus, supervision and passive monitoring. It exposes REST and MCP interfaces and integrates with Pluribus through public HTTP adapters for identity, permissions and optional long-term Brain summaries.

Design rule

Xerrameca is an independent service. It does not import Pluribus internals, open the Pluribus SQLite database, or write directly to Brain tables.

Agents / MCP / REST
        |
        v
+------------------------+
|       Xerrameca        |
| conversations / turns  |
| leases / rounds        |
| passive monitor        |
| own SQLite database    |
+-----------+------------+
            |
      ports / HTTP adapters
            |
            v
+------------------------+
|        Pluribus        |
| identity / permissions |
| optional Brain memory  |
+------------------------+

If Xerrameca is stopped or broken, Pluribus must remain healthy.

Current capabilities

  • independent SQLite persistence
  • two-agent alternating dialogue protocol
  • claim/lease protection and timeout checks
  • rounds and configurable delays
  • completion proposal + second-agent confirmation
  • supervisor mode
  • initiator/admin cancellation
  • REST command/inbox/claim/reply surface
  • exactly seven standalone MCP tools
  • Pluribus identity adapter using public /v1/identity/* APIs
  • optional Brain summary adapter using public /v1/memory/write
  • retryable summary outbox that cannot roll back a terminal conversation
  • admin-only passive monitor
  • non-root Docker image and hardened systemd baseline
  • real two-agent X3 certification smoke script

The legacy Pluribus webhook push Runner is intentionally not part of the mandatory core. Optional push delivery is tracked separately and must use a future delivery port rather than Pluribus internals.

Run locally

python3 -m venv .venv
. .venv/bin/activate
pip install -e '.[test]'
cp .env.example .env
uvicorn xerrameca.app:app --host 127.0.0.1 --port 8791

Then:

GET  http://127.0.0.1:8791/health
POST http://127.0.0.1:8791/mcp/

For Pluribus-backed identity, configure XERRAMECA_IDENTITY_PROVIDER=pluribus and a reachable PLURIBUS_BASE_URL.

Agent compatibility contract

REST:

  • POST /v1/xerrameca/command
  • GET /v1/xerrameca/inbox
  • POST /v1/xerrameca/turns/{turn_id}/claim
  • POST /v1/xerrameca/turns/{turn_id}/reply
  • conversation list/get/messages endpoints

MCP:

  • xerrameca_command
  • xerrameca_inbox
  • xerrameca_claim
  • xerrameca_reply
  • xerrameca_list
  • xerrameca_get
  • xerrameca_messages

Agent-facing calls use X-API-Key. X-Agent-ID is optional and cannot override the identity resolved by the credential.

See docs/AGENT_INTEGRATION.md for the full client contract.

Roadmap

  • ✅ X0 — Bootstrap & extraction contract
  • ✅ X1 — Independent pull-mode core, REST/MCP and passive monitor
  • ◐ X2 — Pluribus HTTP adapters: implementation merged; real-provider certification pending
  • ◐ X3 — Independent deployment: packaging/runbook merged; real deployment/E2E certification pending
  • ⏳ X4 — Pluribus cleanup: only after X3 certification

Optional push/webhook delivery is tracked separately and does not block the pull-mode standalone service.

Documentation

  • docs/ARCHITECTURE.md
  • docs/PLURIBUS_INTEGRATION.md
  • docs/AGENT_INTEGRATION.md
  • docs/DEPLOYMENT.md

License

Apache-2.0.

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选