Factsets MCP Server

io.github.joshua-auchincloss由 GitHub 账号 joshua-auchincloss 发布v0.2.0更新于 Nov 3, 2026

An MCP server that provides a local, persistent, self-maintaining knowledge base for AI agents

已验证STDIO仅桌面知识与记忆开发者工具

概览

AI 生成的概览

为 AI 代理提供基于本地 SQLite 的持久知识库，保存事实、缓存资源、技能文档与执行日志，跨会话保留上下文。

可以做什么
  • 添加、搜索、验证、更新、删除与恢复带标签的事实，并记录来源
  • 注册资源并缓存快照，支持搜索、刷新快照与过期检测
  • 创建、搜索、关联与同步 Markdown 技能文档，查看技能依赖关系
  • 记录并查询命令、测试与构建的执行日志
  • 管理标签、配置项与用户偏好
  • 按标签构建上下文，并获取安装、维护与刷新指南
适用场景
适合希望助手在多次会话之间记住项目知识、为外部文件与 API 保留带过期跟踪的笔记，并沉淀可复用操作技能的场景。长期项目反复重建上下文时尤其有用。
运行要求
以本地 stdio 进程运行，通过 npx、bunx 或 pnpm dlx 启动，需要 Node.js 或 Bun 运行时。使用本地 SQLite 数据库，可通过数据库路径参数指定。未声明需要账号、API 密钥或环境变量。建议执行一次性的安装提示或工具调用以接入项目。
安装前请注意
它会写入本地数据库和项目文件：安装流程可能创建或修改 AGENTS.md、配置技能目录并迁移已有技能，技能文件还会被监视。事实、资源与执行日志会长期保存在本地，应留意助手写入的内容，并在破坏性操作前用导出命令备份。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Factsets MCP Server，将其添加到工作区。
  2. 为需要使用其工具的对话启用该服务。

Desktop only，通过 STDIO。 STDIO 服务会启动本地进程，因此需要 SourceWeft 桌面宿主。

其他 MCP 客户端

参照 仓库 中的启动说明。

README

Factsets

Package[NPM Downloads] [NPM Version]
CI/CD[Release] [Tests] [codecov]

A self-maintaining knowledge base for AI agents, exposed via the Model Context Protocol (MCP). Manages facts (atomic knowledge), resources (cached external content), skills (procedural markdown), and execution logs (command history) using SQLite.

Features

  • Persistent Context - Knowledge survives across sessions
  • Self-Maintaining - Staleness detection with refresh instructions
  • Tag-Based Organization - Flexible categorization and retrieval
  • Skill Documents - Markdown files for procedural knowledge
  • User Preferences - Configurable output style and agent behavior
  • MCP Protocol - Standard interface for AI tool/prompt definitions

Installation

bash
npm install --global factsetspnpm install --global factsetsbun install --global factsets

Quick Start

As MCP Server

Add to your MCP client configuration (Claude Desktop, GitHub Copilot, Cursor, etc.), using bunx, npx or pnpm dlx accordingly:

json
{  "mcpServers": {    "factsets": {      "command": "bunx",      "args": ["factsets", "mcp-server"]    }  }}

Or run directly:

bash
bunx factsets mcp-server

First-Time Setup

After adding Factsets to your MCP client, run the setup prompt to integrate it into your project:

In a supported IDE: Type /mcp.factsets.setup in the chat to run the guided setup

In other clients: Call the get_setup_guide tool or use the setup prompt

The setup guide will:

  • Analyze your project structure and establish baseline facts
  • Configure the skills directory for your AI client
  • Create or update AGENTS.md with Factsets instructions
  • Migrate any existing skills with Factsets integration
  • Register key configuration files as resources

This one-time setup ensures agents have full context on every future interaction.

CLI Commands

bash
# Start MCP server (default command - auto-watches skill files and seeds starter content)bunx factsets [--database-url <path>] [--client <type>]
# Explicit mcp-server command (same as above)bunx factsets mcp-server [--database-url <path>] [--client <type>]
# Start without file watchingbunx factsets --no-watch-skills
# Start without seeding starter contentbunx factsets --no-seed
# Run file watcher standalonebunx factsets watch-files [--database-url <path>]
# Run background maintenance workerbunx factsets worker [--database-url <path>]
# Export database to JSONbunx factsets dump backup.json
# Restore database from JSONbunx factsets restore backup.json

The --client flag configures where skill files are stored (e.g., github-copilot -> .github/prompts/skills/). If you want to change clients / your skill directory, do so through your agent which will migrate skills for you.

See Configuration Guide for all options.

Core Concepts

ConceptDescription
FactsAtomic knowledge units (1-3 sentences), tagged and timestamped
ResourcesExternal content (files, URLs, APIs) with cached snapshots and retrieval methods
SkillsMarkdown documents for procedural knowledge, stored on filesystem
Execution LogsCommand history with success/failure tracking for skill validation
TagsFlexible categorization for all content types

MCP Tools

Facts

ToolDescription
submit_factsAdd facts with tags and source tracking
search_factsQuery facts by tags, content, or filters
verify_factsMark facts as verified by ID
verify_facts_by_tagsBulk verify facts by tags
update_factUpdate fact content, metadata, or tags
delete_factsRemove facts by criteria
restore_factsRestore soft-deleted facts

Resources

ToolDescription
add_resourcesRegister resources with retrieval methods
search_resourcesFind resources by tags, type, or URI
get_resourcesGet resources by ID or URI with freshness
update_resource_snapshotUpdate cached content for single resource
update_resource_snapshotsBulk update cached content
update_resourceUpdate resource metadata (not content)
delete_resourcesRemove resources
restore_resourcesRestore soft-deleted resources

Skills

ToolDescription
create_skillCreate markdown skill document
update_skillUpdate skill metadata/references
search_skillsFind skills by tags or query
get_skillsGet skills by name with content
link_skillLink skill to facts/resources/skills
sync_skillSync skill after file edit
delete_skillsRemove skills
get_dependency_graphGet skill dependency tree
restore_skillsRestore soft-deleted skills

Execution Logs

ToolDescription
submit_execution_logsRecord command/test/build executions
search_execution_logsFind executions by query, tags, success
get_execution_logGet execution details by ID

Tags

ToolDescription
create_tagsCreate organizational tags
list_tagsList tags with usage counts
update_tagsUpdate tag descriptions
prune_orphan_tagsClean up unused orphan tags

Configuration

ToolDescription
get_configGet a configuration value by key
set_configSet a configuration value
delete_configDelete a configuration value
list_configList all configuration with schema
get_config_schemaGet available options with descriptions

User Preferences

ToolDescription
get_preference_promptGet natural language preference prompt
get_user_preferencesGet structured preference data
infer_preferenceUpdate preference from user behavior
reset_preferencesReset preferences to defaults

Maintenance

ToolDescription
check_staleFind stale resources and dependencies
mark_resources_refreshedMark resources as current

Context & Guides

ToolDescription
get_knowledge_contextBuild context from tags (facts/resources/skills)
build_skill_contextGet skill with formatted content and refs
get_maintenance_reportGenerate staleness/maintenance report
get_refresh_guideGet instructions for refreshing a resource
get_agent_guideGet the agent workflow guide (call first)
get_concept_guideGet conceptual overview and design philosophy
get_config_guideGet configuration guide with all options

MCP Prompts

PromptDescription
setupGuided setup for new project integration
user_preferencesGet user preferences for output formatting
knowledge_contextBuild context from tags
recall_skillGet skill with references
maintenance_reportStaleness summary
refresh_guideInstructions to refresh a resource
agent_guideAgent workflow guide (call first)
conceptConceptual overview and philosophy
configConfiguration guide with all options

Documentation

Development

bash
# Run testsbun test
# Run full e2e (tests + build + dry run)bun e2e
# Build distributionbun dist
# Format codebun format
# Lintbun lint
# Generate database migrationsbun migrations
# Inspect MCP server with inspectorbun inspect

来源：README.md，提交 1cf149c

工具

0
SourceWeft 尚未在沙箱中启动过该软件包，因此其工具未知。

版本历史

1
  1. v0.2.0最新Sep 16, 2026

评分

已满足适用检查项的 50%

该评分衡量此收录条目的完整程度，依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价，也不表示服务器安全或当前可用：可连通性是带日期的检查结果，评分也不包含安装量或用户评价。

  • 近 30 天内可连通25 分

    请保持端点能够响应 MCP 握手；SourceWeft 每 30 天探测一次。

  • 已声明认证方式10 分

    请在注册中心条目中声明凭据请求头，或保持端点能够响应探测，以便判定其认证方式。

  • 已列出工具15 分

    请让服务器在无凭据时响应 tools/list，或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分