Brain Framework (bf) turns a folder of plain Markdown notes and collected JSON records into a local, verifiable knowledge graph for you and any coding agent.
- Plain files you own. Notes (
projects/,concepts/,actions/), records (memories/), and config (bf.yaml,evals/), accessible throughbf, local stdio MCP, Agent Skills, andAGENTS.md. - Gather, connect, act offline. Scheduled or streaming
jobs:collect or ingest records (bf collect,bf ingest), declarativemap:normalizes fields, BM25 + graph links trace every claim, and prompt jobs wake harnesses only when events or gates fire. - No lock-in or built-in LLM. BF runs zero models, accounts, or telemetry; search and read stay 100% offline.
Try it
Requires Linux or macOS and uv — no model or cloud account needed:
uv tool install --python 3.14 brain-framework
bf init ~/brain
cd ~/brain
bf guide # Offline documentation inside the CLI
bf validate
bf eval
Explore the four-tool walkthrough (Workspace, Jira, GitHub, and Google Cloud fixtures) or read Getting started.
How it works
- Notes & graph (Authoring): OKF Markdown notes with explicit
relations:andattributes:, plus trackedACTION.mdworkflows (bf action new|status). - Offline retrieval (Retrieval): Disposable
.bf/cache.sqlite3powersbf search,bf read, review pages,bf export,bf eval, andbf mcp. - Agents & skills (Getting started): Install the
/bfskill (bf skill install --global) and register brains (bf register . --default); optional host hooks surface brain state automatically. - Sensors, ingest & routines (Jobs): Copy stdlib scripts from sensors and routines, pipe external JSON via
bf ingest, and normalize fields with declarativemap:. - Prompts, channels & watch (Assist): Event-driven prompt jobs (
record.*,note.changed,message.received,job.failed), DM allowlists ({{lines:...}}), durable channel queues, failure incidents, andbf watch(demo).
Good to know
- Local-first and private. Retrieval runs offline;
.bf/locks, caches, and durable queues (runtime.sqlite3) never touch the network (Operations). - Harnesses reason; BF verifies. BF invokes your configured CLI harness only on explicit runs, schedules, or event/gate matches, and validates every note, record, and
bf.yaml(version: 11) (Reference). - Git-native collaboration. Track notes,
bf.yaml,jobs/, andevals/in Git while ignoringmemories/,actions/,logs/, and.bf/.
Documentation · Reference · Issues · Contribute · Governance · MIT License · Third-party notices
Metadata
Release files for brain-framework 25.0.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| brain_framework-25.0.0.tar.gz | 227.2 kB | With provenance | Details for brain_framework-25.0.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| brain_framework-25.0.0-py3-none-any.whl | Python 3 | none | any | Details for brain_framework-25.0.0-py3-none-any.whl |
Total release size: 473.6 kB
Release files / brain_framework-25.0.0.tar.gz
| Download URL | brain_framework-25.0.0.tar.gz |
|---|---|
| Size | 227.2 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
3a174ecb0995b4425915640c411d38a04bb0cdb0242da0de1f8a54dcd3cf46d6
|
|
BLAKE2b-256 checksum How to use checksums |
85e9111c58074c7f2dc9932974a83f4f10ab8b55096dc8f9eb4241c612d81050
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.16
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Nov 14, 2026.
Transparency logRelease files / brain_framework-25.0.0-py3-none-any.whl
| Download URL | brain_framework-25.0.0-py3-none-any.whl |
|---|---|
| Size | 246.4 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
ad8e2dfc7f01f455b09cf6737bf694f070c8372b6b1da91e0681f05059870170
|
|
BLAKE2b-256 checksum How to use checksums |
068e7da8d3faf94686eda98461704300c664d0857bc29c62103698f255460d33
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.16
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Nov 14, 2026.
Transparency logSecurity for brain-framework 25.0.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
Full provenance
Both files in this release have been uploaded with provenance metadata.
Attestation for brain_framework-25.0.0.tar.gz
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Nov 14, 2026.
Transparency logA note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.