English · 繁體中文
Muninn is a personal assistant that runs on your own computer, and that you talk to in one conversation, on your phone or in a browser. It keeps track of what you are working towards, notices when you drift from the plan, and brings up only what matters, with what to do next. Whatever it does on its own can be undone; anything that cannot be taken back waits for your yes.
Install
On macOS or Linux, this installs Muninn, starts it in the background and opens its setup page:
curl -fsSL https://raw.githubusercontent.com/imaknas/muninn/main/install.sh | sh
Or with uv:
uv tool install muninn-ai
muninn init && muninn serve
On a NAS or a server: docker compose up -d (see compose.yaml). More in docs/install.md.
Getting started
- On the setup page, pick a model and paste its key: Claude, any service with an OpenAI-style API, or a local model with Ollama.
- Start talking to it, right there in the browser.
- For your phone, open Settings → Pair a phone. With Tailscale on the computer, Muninn gives itself an HTTPS address on your tailnet; scan the code.
- The rest is on the settings page too: the folders it reads activity from, connecting Google (mail and calendar, step by step), and adding form filling.
muninn upgrade moves to the latest release, muninn check says what in the settings cannot work as written,
muninn backup DIR copies your data, and muninn uninstall takes Muninn off the computer again.
What it does
- One conversation. Tell it what happened, ask it things, answer what it asks. No forms.
- Goals and deadlines, per part of your life. It measures progress from what it observes and nudges you when you fall behind, within daily limits and quiet hours.
- A morning brief, an evening summary and a weekly reflection.
- Mail and calendar. Google: it reads mail, saves reply drafts and blocks time in a calendar of its own. Microsoft 365: it reads. It never sends mail or answers an invitation for you.
- Job applications. It prepares each one from the posting and your documents; with the
autofillextra installed it also fills in Greenhouse forms, for you to send with one tap. - Your coding projects. A project can hand it to-dos and questions only you can settle (Principal Handoff), and with a session runner (such as agent-hub) a conversation can be lent to a project's own session.
- English or Traditional Chinese, throughout.
Privacy
Muninn runs on your machine, and what it knows stays in its data/ folder. One setting decides where models run:
cloud sends every request to the cloud model you chose; privacy keeps work and anything unattributed on a local
model, and gives the cloud only whitelisted fields. You can cap what cloud models cost per day, and every device has to
be paired. Details in docs/privacy.md; vulnerabilities go to SECURITY.md.
Documentation
| Install, upgrade, Docker, HTTPS for your phone | docs/install.md |
| Domains, goals, what it observes, what it does for you | docs/how-it-works.md |
| Where models run and what leaves the machine | docs/privacy.md |
| Interfaces for other systems | src/muninn/contracts |
| Developing, testing, releasing | CONTRIBUTING.md |
| What changed in each release | CHANGELOG.md |
License
Metadata
Release files for muninn-ai 0.8.0
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 |
|---|---|---|---|---|
| muninn_ai-0.8.0.tar.gz | 710.3 kB | With provenance | Details for muninn_ai-0.8.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| muninn_ai-0.8.0-py3-none-any.whl | Python 3 | none | any | Details for muninn_ai-0.8.0-py3-none-any.whl |
Total release size: 1.6 MB
Release files / muninn_ai-0.8.0.tar.gz
| Download URL | muninn_ai-0.8.0.tar.gz |
|---|---|
| Size | 710.3 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
92abd124e3329b517ec6ccd6712f2747e8a025eb5a12809a4a34dfb2f58a6749
|
|
BLAKE2b-256 checksum How to use checksums |
c9b8929869b89283b7c015ba4023cca7fa3bf472a6c3b7ed38ecbeeae0fbb7a8
|
| 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 2, 2026.
Transparency logRelease files / muninn_ai-0.8.0-py3-none-any.whl
| Download URL | muninn_ai-0.8.0-py3-none-any.whl |
|---|---|
| Size | 884.4 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
7c60854c7f94b6104a6b360d55f2326a51566f6b3b61cd0dce6fecfaf9b9fa28
|
|
BLAKE2b-256 checksum How to use checksums |
1d5a87f09d3f03bd53a573485bd12ab7c406dead6dc332f38d14ae3445bb5bf4
|
| 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 2, 2026.
Transparency logSecurity for muninn-ai 0.8.0
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
All 2 files in this release have been uploaded with provenance metadata.
Attestation for muninn_ai-0.8.0.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 2, 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.