📁文件系统 MCP 推荐

Local-first memory + judgment layer for AI coding agents — warns before repeating failed fixes.

Read and resolve human feedback pinned on your live prototype, straight from your coding agent.

Post, delete & read on X (Twitter) using your own logged-in web session — no paid API.

MCP server for Callimachus: local index and search of your AI coding-agent threads

Architecture governance for AI agents: query the live model, get target patterns, check files.

Get today's WriteFlow daily writing prompt and weekly £10 writing contest info (writeflow.ink).

Local-first TDLib Telegram MCP server with 110 tools, account isolation and production guardrails.

Public read-only MCP over the complete I Ching hexagram dataset. No divination.

Delegate mechanical text tasks to a local OpenAI-compatible LLM to save Claude quota.

MCP (Model Context Protocol) server for Appwrite

The agent never writes SQL: queries compile from a semantic layer you control, PII refused first.

Live context MCP server: resolves git, services, and tests into agent context. Local-first, MIT.

Securely give LLMs controlled access to real-world operations inside your Files.com environment

Local coordination hub for coding agents: durable shared memory, tasks, and handoffs. Desktop app.

Agent-first US business trust registry with neutral Trust Cards and local search.

Deploy + manage apps, domains, VPS & crypto balance on Impreza Host - no-KYC, offshore, BTC/XMR.

File-first, local-first MCP memory for AI coding assistants: Markdown + YAML, no vector DB.

Independent MCP server for the TC39 specs (ECMA-262 + ECMA-402): clauses, search, diffs, history.

Publish HTML, files, or a URL to a permanent public URL, then update it — from any MCP agent.

Your org's AI agents, tasks, runs, search, and brain files as MCP tools and resources.

Create App Store screenshots, icons, ASO copy, localization, and revisions via hosted MCP.

Real signed-browser automation for AI agents, over a local stdio MCP server.

Your project's decision history for agents: decisions, dead-ends, constraints - queryable, local.

Every NinjaOne report, plus a local store that answers fleet-wide questions no single API call can

RoastPilot controls coffee roasting sessions through local stdio MCP tools and exports logs.

Keyless web access for AI agents: read bot-walled pages as Markdown, search, grab assets

MuseScore for Claude — search sheet music and read score metadata via your signed-in browser

Canvas LMS (Instructure) for Claude — courses, grades, assignments, planner, files

Read-only Workday for Claude — fetch tasks, pay, benefits via your signed-in session

Cross-session memory for Claude over MCP — prioritised briefs, supersession, cross-project recall.

Local-first miniature-paint inventory and cross-brand color matching for AI agents.

Local-first task, time, expense, invoice, and planning tools.

Connect MCP-capable AI agents to local Granoflow tasks, reviews, cards, imports, and work memory.

Extract structured, schema-validated data from PDFs, scans, images, spreadsheets, and forms.

Local-first, multi-provider tool that captures LLM API spend and exposes it to coding agents via MCP

Query any database in plain English or SQL through a guardrailed, read-only data layer.

Local-first social MCP: an agent drafts and schedules posts; you approve before anything publishes.

Audits any website for AI agent readiness and safety, scoring prompt injection risk and llms.txt.

Convert PDF, DOCX, XLSX, CSV, JSON into clean, AI-ready Markdown via the PackForAI API.

Read-only MCP server: query_metrics over local metric time-series (Quellgeist triage).

Read-only MCP server: query_logs over a structured JSONL incident log (Quellgeist triage).

Read-only MCP server: get_recent_commits over a local deploy log (Quellgeist triage).

Phone-first local dev MCP for Claude Code and Codex: install Yaver, pair mobile, run, build, deploy.

Bug translation for vibe coders: plain-English diagnosis + paste-ready fix in your AI coding agent

Git-native, review-gated knowledge base for LLM agents. Cited retrieval, audited writes.

Upload a file and get a public shareable URL via API or MCP.

Read-only MCP server for the WebAssembly spec: instructions, types, sections, search, proposals.

Runnable, versioned lists agents can search, read, create and update. Git under the hood.