Bailey
An open-source AI assistant that runs ready-made solutions: on your own machine, with your own AI key.
Early days. This is release 0.0.1. It installs one command,
bailey doctor, which checks that your computer has what the assistant will need. The assistant itself arrives in 0.1.
What Bailey is for
Getting AI to do one real job reliably is still a project. Someone writes the instructions, connects the tools, supplies the documents and the keys, checks that it works, and keeps it working when something changes.
Bailey runs solutions instead: complete, versioned packages for one job, such as answering customer emails, qualifying leads or writing a weekly report. A solution carries its instructions, skills, tools, knowledge and small apps. Bailey installs it in one step and asks only for the keys it needs.
- Your machine, your key. It runs locally and opens in your browser, with the AI provider you choose: Anthropic, OpenAI, or any OpenAI-compatible endpoint, local models included.
- It asks before it acts. Every tool is either read or write, and a write tool asks before it runs: once, always for this assistant, or no.
- Your documents stay yours. Files you add are indexed on your computer, and answers cite the passage they came from.
- Open formats. Skills are Agent Skills, tools are MCP servers, apps speak Postern, and a solution is an Agent Plugins folder, so its parts also load in other clients.
- No telemetry.
Solutions can be free, or bought on Sigrix, where builders sell what they make for Bailey.
Check your machine
uvx bailey doctor
pipx run bailey doctor works as well. It checks, and says how to fix what is missing:
| Check | Why Bailey needs it | Required |
|---|---|---|
| Python 3.11 or newer | Bailey itself | yes |
| SQLite with full-text search (FTS5) | searching your documents by keyword | yes |
| A loopback address | the app's local server, which only this computer can reach | yes |
| SQLite extension loading | the vector index, which finds passages by meaning | no |
| uv | tools published as Python packages, which start with uvx |
no |
| Node.js | tools published as npm packages, which start with npx |
no |
It exits 1 when a required check fails, and bailey doctor --json prints the same answers for a script. It reads your machine only: no network call, no file written.
What comes next
- 0.1: the app. Chat, assistants, skills loaded when needed, tools with approval, knowledge from your own documents with citations, apps, and installing a solution from a file.
- 0.2: the store. Sign in, browse, install and update solutions, and publish your own.
Development
pip install -e ".[dev]"
ruff check . && ruff format --check . && pytest
Licence
Apache-2.0. The Sigrix name and logo are not covered by the licence; see NOTICE.
Metadata
Release files for bailey 0.0.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| bailey-0.0.1.tar.gz | 12.9 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| bailey-0.0.1-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 24.4 kB
Release files / bailey-0.0.1.tar.gz
| Download URL | bailey-0.0.1.tar.gz |
|---|---|
| Size | 12.9 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
fb8999ea6935958fe2db5356f0099e8e1208667829d32ae13809778cece9f919
|
|
BLAKE2b-256 checksum How to use checksums |
7931144d3e3432e5a3247f99b741562f4cf1afc3faaca98b198ecea536eaa637
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
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 Sep 29, 2026.
Transparency logRelease files / bailey-0.0.1-py3-none-any.whl
| Download URL | bailey-0.0.1-py3-none-any.whl |
|---|---|
| Size | 11.5 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
1b829b372b529363f901a54bdce77eddd04b32b00a733d3ca1b3b6ad7e8270be
|
|
BLAKE2b-256 checksum How to use checksums |
e4e237f37324cae737b65c578908c3c1309b6f75785e7b7a22d47b8f724e0e4d
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
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 Sep 29, 2026.
Transparency log