Synthadoc
A CLI engine that compiles documents into a local structured wiki: a transparent alternative to plain RAG, reachable by an agent over MCP
Medium risk
We rate an entry medium when the tool runs code, makes network calls or reads project files. Check what exactly it does before installing.
Why this level
- The engine reads your documents and writes wiki files
- It runs a local server and may use a local model
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/synthadocDetects the agents on your machine, checks the risk and pins the version.
Other ways to install
Install the tool
pipx install synthadocWithout pipx, pip install synthadoc works too.
Install the engine: pip install synthadoc, compile the wiki with CLI commands, then connect it to the agent: claude mcp add --transport sse synthadoc-my-wiki http://127.0.0.1:7070/mcp/sse.
Other ways from the author
pip install synthadocAfter install, compile wikis with the synthadoc CLI commands.
This is third-party code. Review the repository files before installing.
What it does
Synthadoc takes source documents and compiles them into a structured wiki in human-readable markdown files on your machine. Unlike plain RAG the result is transparent: pages can be opened and read directly, not only queried through vectors. The engine installs as a pip package and is driven by CLI commands, with a plugin system for ready knowledge sets. A compiled wiki can be exposed to an agent over MCP through a local address so it can search and answer from it.
Who it is for. For analysts, technical writers and developers who need a transparent knowledge base from their own documents.
Good fit when
- You need to compile a structured wiki from a document set
- You want a transparent RAG alternative with readable pages
- You need to give an agent local access to the knowledge base over MCP
Not a fit when
- You need a hosted knowledge-base service without local setup
- You have few documents and need no structure
Example request
Compile these documents into a structured wiki and connect it to the agent for answersLimitations
Synthadoc is a local engine: it must be installed via pip and run on your side, and compilation may need dependency setup and a local model. The AGPL-3.0 license imposes conditions on distribution. MCP access works through a local server you must keep running.
How to disable. Remove the package with pip uninstall synthadoc and delete the compiled wiki directories and the MCP server entry from your client config.
Security check
- The engine reads your documents and writes wiki files
- It runs a local server and may use a local model
README in short
The README describes Synthadoc as an open-source engine that compiles documents into a structured local wiki as readable markdown files. It is a transparent alternative to plain RAG that can be maintained and improved by hand. The engine installs via pip, is driven by a CLI, supports knowledge-set plugins and exposes the wiki to an agent over MCP through a local address. It has an Obsidian plugin. AGPL-3.0 licensed.
FAQ
How is this different from plain RAG?
The result is readable markdown pages you can open and check, not only a hidden vector index.
How does the agent access the wiki?
Through a local MCP server on a loopback address that you keep running.
Related
166 skills for scientific work: bioinformatics, cheminformatics, clinical data, geospatial analysis and 100+ databases
Google's open-source MCP server for databases: ready tools for Postgres, MySQL, BigQuery, Spanner and more, plus custom tools in tools.yaml
Official Hugging Face skills: Hub operations via the hf CLI, datasets, model training, Spaces, evals and deployment
A visualization language for agents and an MCP server: neat charts from a simple semantic spec