SagaSmith CoC MCP

SagaSmith CoC MCP

Local MCP server for Call of Cthulhu 7e, managing campaigns, character knowledge, and rule resolutions (d100, SAN, combat) with secure state isolation and keeper-controlled access.

Category
Visit Server

README

SagaSmith CoC MCP

中文 · English · 平台总览

SagaSmithAI 的 Call of Cthulhu 7e 本地 MCP 服务。 它把 sagasmith-core 的战役、分支记忆、角色知识、Snapshot 与模组能力,以及 sagasmith-coc 的 d100、SAN、战斗和追逐判定,收敛到一个可被任何 MCP Agent 使用的服务端边界。

为什么是独立 MCP

  • 状态归 MCP 所有:SQLite 与导入产物都位于 .sagasmith-coc-mcp/,不依赖某个 Agent 的工作目录。
  • Exposure 在服务端:每个原生 MCP session 独立维护已加载工具组、TTL 与 Lobby/Play/Combat 阶段,不要求 Agent 复制工具分类。
  • 角色知识真正隔离:PC、NPC、怪物的 belief/rumor/false belief 分别按 actor 与 branch 保存;玩家只能读取或修改被授权角色。
  • 判定与写状态分离:coc_resolve 返回规则结算;角色卡、SAN、HP 或战役状态必须通过显式写工具提交,避免“调用检查就偷偷改卡”。
  • Keeper 信息不外泄:玩家的模组索引与搜索只返回 visibility=player 的 handout 场景。
flowchart LR
    A[Agent / CoC Skills] --> E[Session Exposure]
    E --> M[SagaSmith CoC MCP]
    M --> C[sagasmith-core]
    M --> R[sagasmith-coc]
    C --> D[(SQLite · branches · actor knowledge)]

工具阶段

阶段 能力组 用途
Lobby lobby.bootstrap 建战役、成员与角色授权
Lobby lobby.characters 车卡、NPC/怪物、角色卡维护
Lobby lobby.modules 导入模组、检查 scene index
Lobby lobby.continuity 记忆、角色知识、Snapshot、Skills
Play play.investigation 场景推进、调查判定、SAN、追逐与连续性
Combat combat.resolve 近战/远程判定与 Keeper 显式状态写入

Agent 首次只会看到目录级核心工具。标准流程是:

exposure_open → exposure_search → exposure_inspect → exposure_load → domain tool

不支持动态刷新工具列表的 Host 可以使用 exposure_call,权限与 TTL 仍由 MCP 服务端校验。

快速开始

pip install -e "../sagasmith-core[documents]"
pip install -e ../sagasmith-coc
pip install -e .
sagasmith-coc-mcp

默认状态目录:./.sagasmith-coc-mcp/。可用 SAGASMITH_COC_MCP_HOME 指定路径,并用 SAGASMITH_COC_SKILLS_DIR、SAGASMITH_MODULEGEN_SKILLS_DIR 接入 Skills 生态。

安全边界

  • 创建战役时创建 owner membership;新增玩家需 Keeper 显式 grant campaign 与 actor。
  • owner/DM 可管理全部角色;玩家只能访问明确授权的 actor private state。
  • Exposure 同时绑定 MCP session、principal、campaign 与阶段,不能跨会话或跨战役复用。
  • combat.active=true 时进入 Combat;非 Keeper 不能直接写角色卡,以免绕过判定与行动边界。
  • 商业规则书和模组不随仓库发布;仅导入你有权使用的内容。

开发

pip install -e ".[dev]"
pytest
ruff check .

代码使用 MIT License。Call of Cthulhu 及相关商业内容归其权利人所有。

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured