About KiwiFS
KiwiFS is a Markdown filesystem for agents and teams. Its premise is that raw Markdown files are just files, with no search, versioning or structure, so it makes them writable, searchable, queryable and versioned while keeping files as the source of truth and every other index rebuildable. It ships as one binary with zero configuration, or runs in Docker.
It offers 62 MCP tools for clients such as Claude and Cursor, a web UI with a block editor, wiki links, backlinks and a knowledge graph embedded in the binary, full-text search using SQLite FTS5 with pluggable vector search backends, and Git versioning where each write is an atomic commit with blame, diff and point-in-time restore. DQL, a SQL-like query language over frontmatter, supports TABLE, LIST, COUNT, WHERE, SORT and GROUP BY. Files are reachable over REST, MCP, NFS, S3, WebDAV and FUSE, and 19 importers bring in data from Postgres, MySQL, MongoDB, Notion, CSV and Obsidian.
Further features include content health checks for stale pages, broken links and orphans, multiple isolated spaces, and webhooks for write and delete events. KiwiFS is written in Go and self-hosted; its license is listed as Other.
Key features
- Markdown files with Git versioning
- 62 MCP tools for AI clients
- Full-text and vector search
- DQL queries over frontmatter
- Access via REST, S3, NFS, WebDAV and FUSE
- Importers for Notion, Obsidian and databases
Good fit for
- Shared memory and knowledge base for AI agents
- Searchable, versioned team wiki in Markdown
KiwiFS: questions and answers
- What is KiwiFS used for?
- KiwiFS is a single-binary Markdown filesystem for AI agents and teams that adds search, Git versioning, queries and 62 MCP tools to plain files. It is a good fit for shared memory and knowledge base for AI agents; and searchable, versioned team wiki in Markdown.
- Is KiwiFS open source?
- Yes. KiwiFS is open source under a custom licence. Its source code is on GitHub at kiwifs/kiwifs and is written mainly in Go.
- Is KiwiFS free?
- Yes. KiwiFS is open source, so the software itself is free to use under the terms of its own licence.
- Can I self-host KiwiFS?
- Yes. KiwiFS can be self-hosted on your own server or infrastructure.
- What are some alternatives to KiwiFS?
- Similar open-source tools in the AI Infrastructure category include Yuxi, RAGFlow and OpenConnector. SaaS products in the same category include IgnitionRAG, Lobstack AI and Surf Data.
- Is KiwiFS actively maintained?
- Yes. The most recent commit to KiwiFS was on 11 November 2026, and the latest release is v0.19.90, published on 11 November 2026. The project has 877 stars on GitHub, with 157 added in the last 30 days.
Open-source alternatives to KiwiFS
See all
Yuxi
AI Infrastructure
可私有部署的多租户知识智能体平台,统一知识库、知识图谱、多智能体执行、MCP/Skills、沙盒与权限管理。
MIT★ 7.7k
RAGFlow
AI Infrastructure
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cu
Apache-2.0vs Vectara★ 94k
OpenConnector
AI Infrastructure
Connect Gmail, GitHub, Notion, and 1500+ apps through OOMOL, then use those connections in
Apache-2.0★ 6.2k
MCP Toolbox
AI Infrastructure
MCP Toolbox for Databases is an open source MCP server for databases.
Apache-2.0vs Connect★ 17k
AnythingMCP
AI Infrastructure
Open-source, self-hosted MCP gateway: turn any REST/OpenAPI, SOAP, GraphQL, OData or SQL A
AGPL-3.0★ 776
Inkeep Agents
AI Infrastructure
Create AI Agents in a No-Code Visual Builder or TypeScript SDK with full 2-way sync. For s
OSS★ 1.5k
SaaS alternatives to KiwiFS
See all
IgnitionRAG
AI Infrastructure
RAG platform covering document ingestion, search, agents, evaluation and observability for AI assistants
SaaS
Lobstack AI
AI Infrastructure
Infrastructure that gives each AI agent a dedicated VM with persistent memory, tools and 117+ integrations
SaaS
Surf Data
AI Infrastructure
Platform that exposes database data to AI agents through MCP, with built-in LGPD compliance
SaaS
AgentGov
AI Infrastructure
Open source platform for AI agent tracing and EU AI Act compliance with audit trails and risk documentation
SaaS
Amazon Bedrock
AI Infrastructure
Managed AWS service for building generative AI apps with models from several providers
SaaS
AnyRouter
AI Infrastructure
Unified LLM gateway that routes coding agents and MCP servers through a single endpoint
SaaS

