mcp-server

pharos – Fast, Local‑First RAG for Your Team’s Docs

Name: pharos – Fast, Local‑First RAG for Your Team’s Docs
Author: Laurent00TT

pharos lets founders instantly query multi‑format document libraries with hybrid search and granular ACLs, using a self‑hosted Python API and MCP/HTTP endpoints.

From the repo · “Pharos — local-first agentic RAG for your team's document library: multi-format ingest, hybrid retrieval, enterprise ACL, dual HTTP + MCP exits.”

241 stars1 forksPythonGuide quality 8/10 · reviewed Oct 22Updated 8/19/2026100% free · open source

What it does

Pharos provides a self-hosted document library with hybrid search and retrieval capabilities for startup founders.

When to use it

•When you need to manage a large collection of documents in multiple formats
•When you require a high degree of control over access and permissions for your document library
•When you want to integrate a document library with other tools and systems using both HTTP and MCP exits

Ready-to-paste prompt

python ingest.py --format pdf --path /path/to/documents && curl http://localhost:8000/search?q=example

Heads up: Make sure you have the correct version of Python installed, as Pharos requires Python 3.8 or later, and also ensure you have the necessary dependencies installed, such as `pyyaml` and `pdfplumber`, which are specified in the `requirements.txt` file

Saves to your device

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

🔑 Needs an API key

Works after you add credentials — the setup agent will ask you for them.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/pharos && curl -fsSL https://workflowstacks.com/api/skills/pharos/claude-skill -o ~/.claude/skills/pharos/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How pharos – Fast, Local‑First RAG for Your Team’s Docs works

Codeflow

Free to inspect

Pharos – Fast, Local‑First RAG for Your Team’s Docs is a medium Python project (~12k lines across 72 code files, plus 20 test files). Setup is light: runs with Docker — add your API keys to .env. Last commit 2 months ago, MIT license, has a test suite.

Size

Medium codebase

~12k lines · 72 code files · ~2 h to skim

Setup

Light setup

Runs with Docker — add your API keys to .env.

Runs on

Python · Docker

Needs API keys (.env)

Python 100%

Where to start reading

1
README.md
Start here — what it does and how to install it
2
pyproject.toml
Dependencies and the commands it exposes
3
.env.example
The API keys and settings you must provide
4
src/chunker/__init__.py
Inside src/ — the main logic begins here

What's in each folder

src/Core code — the actual logic45 files

docs/Documentation45 files

eval/Evaluations & benchmarks22 files

scripts/Helper scripts10 files

config/Configuration1 files

deploy/Deployment / infrastructure1 files

tests/Tests — proof it works23 files

.github/CI / automation (GitHub Actions)1 files

READMEHas testsDocumentedCI checksDocker readyMIT licenseUpdated 2 mo ago

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

🔑 Needs an API key

Works after you add credentials — the setup agent will ask you for them.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/pharos && curl -fsSL https://workflowstacks.com/api/skills/pharos/claude-skill -o ~/.claude/skills/pharos/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

Quick Actions

Details

Creator

Laurent00TT

Language

Python

Do this with a working workflow

Pages we wrote and tested that put a skill like this to work — importable n8n templates, outcome guides and Claude Desktop configs.

MCP config

Context7 (live docs) MCP config

Pull up-to-date library docs into Claude so it codes against the real API.

MCP config

Memory (knowledge graph) MCP config

Persistent memory across chats via a local knowledge graph.

Related skills

More mcp-server tools founders pair with this one.

mcp-server★ 365,609

Rea: Reverse Engineer Apps

Inspect apps down to binary level with Rea. For founders needing code clarity.

mcp-server★ 156,480

open-webui: Self‑hosted AI Chat UI for Faster LLM Integration

Launch a sleek, web‑based AI interface that works with Ollama, OpenAI API, and more—ideal for founders needing a quick, customizable LLM front‑end. 153k+ GitHub stars.

mcp-server★ 133,178

graphify

Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.

mcp-server★ 16,709

mcp-toolbox: Unified Database MCP Server for Fast Scaling

Leverage mcp-toolbox to manage multiple databases through a single MCP server, cutting integration effort for founders building data‑centric apps. Backed by 16k+ GitHub stars.

mcp-server★ 1,941

video-search-and-summarization: Build GPU‑fast vision agents for video AI

Leverage ready‑made architectures to add real‑time video search, summarization and analytics to your product. Ideal for founders building AI‑powered vision services. 1.9k+ GitHub stars

mcp-server★ 3,989

linkedin-mcp-server: Integrate AI Agents

Give AI agents access to LinkedIn data with linkedin-mcp-server, suitable for founders integrating AI tools, backed by 2.9k+ GitHub stars.