mfs-server

The server side of MFS — Multi-source File-like Search: a context engine that turns code, docs, messages, databases, and object stores into one file-like, searchable namespace for AI agents and developers.

mfs-server is the heavy half of MFS. It owns the connectors, the ingest pipeline, the vector index, and the storage — and exposes everything over an HTTP /v1 control plane. The thin mfs CLI and the generated SDKs are just clients of this server.

Install

uv tool install mfs-server                     # core + local ONNX embeddings, Milvus Lite, SQLite
uv tool install "mfs-server[all-connectors]"   # add every connector's SDK
uv tool install "mfs-server[pg,s3,slack]"      # or just the connectors you need

uv tool install puts mfs-server on your PATH in an isolated environment. The defaults run fully offline: local ONNX (BGE) embeddings, Milvus Lite, and SQLite, with no cloud account or API key.

Run

mfs-server setup     # optional: write $MFS_HOME/server.toml (defaults to ~/.mfs)
mfs-server run       # all-in-one: API + inline task processing on 127.0.0.1:13619

For a horizontally scaled deployment, split the roles:

mfs-server api                    # HTTP control plane only
mfs-server worker --concurrency auto   # queue worker(s)

Point the mfs CLI (or an SDK) at the server and you're ready to mfs add a source and mfs search it. See the Quickstart for the full first run.

Optional native acceleration

A few hot paths — the gitignore directory walk, parallel content hashing, linear grep, and tail — have an optional PyO3 extension (mfs-server-rs). The server falls back to a pure-Python implementation when it isn't installed, so it is never required; it only makes large inputs faster. See Deployment for how to build it in.

License

Apache-2.0.

Project links

Data verified by PyPI on
Data provided by the project maintainers, verified at the time the release was uploaded to PyPI.

Key dates

PyPI data
Data sourced directly from PyPI's database.
  • Released:
Newer release available (0.4.6)

1 maintainer

PyPI data
Data sourced directly from PyPI's database.
Avatar for zc277584121 from gravatar.com zc277584121

Credits

Author: Cheney Zhang

GitHub Statistics

Data verified by PyPI on 
The GitHub source repository was provided by the project maintainers and verified by PyPI at the time of upload. Stars, forks, and open issues/PRs are derived from that repository and have not been independently verified.

License expression

Apache-2.0
View SPDX License List

Requires

Python >=3.10

Provides Extra

all-connectors all-providers anthropic bigquery dev embedding-onnx feishu gemini github google hubspot jira local mistral notion ollama pg s3 slack snowflake voyage web

Classifiers

Report project as malware

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

mfs_server-0.4.5.tar.gz (684.2 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

mfs_server-0.4.5-py3-none-any.whl (304.9 kB view details)

Uploaded Python 3

File details

Details for the file mfs_server-0.4.5.tar.gz.

File metadata

  • Download URL: mfs_server-0.4.5.tar.gz
  • Upload date:
  • Size: 684.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for mfs_server-0.4.5.tar.gz
Algorithm Hash digest
SHA256 3f0d10971c02d291ad5feae5531539ce679d7b5f0d383a6099430ad2321a008b
MD5 ad691d5be8832747b34b09967cee6960
BLAKE2b-256 cb2839643357dead356e7e85c0df0a9d0013d5d30b44e3b772c4f48d3aa7e460

See more details on using hashes here.

Provenance

The following attestation bundles were made for mfs_server-0.4.5.tar.gz:

Publisher: publish-pypi.yml on zilliztech/mfs

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file mfs_server-0.4.5-py3-none-any.whl.

File metadata

  • Download URL: mfs_server-0.4.5-py3-none-any.whl
  • Upload date:
  • Size: 304.9 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for mfs_server-0.4.5-py3-none-any.whl
Algorithm Hash digest
SHA256 e5f89a6cd836def4f33547f52cce7e3c95b80f76f56ad8384f4ead4560c1a430
MD5 2b14bf76669087a21f918fc5db6e6f7e
BLAKE2b-256 80d863fdef6928acd342181be48339807296a6c38976fa33bce821cb0e465b3b

See more details on using hashes here.

Provenance

The following attestation bundles were made for mfs_server-0.4.5-py3-none-any.whl:

Publisher: publish-pypi.yml on zilliztech/mfs

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page