Local behavioral context for AI coding assistants. Zero accounts, zero cloud.
Project description
🧠 Humsana Local
Your AI coding assistant doesn't know if you're in deep focus or exhausted. Now it does.
A tiny local server that gives Continue.dev (and other AI tools) real-time awareness of your focus state. No cloud, no accounts, no tracking.
🚀 Install
pip install humsana-local
▶️ Run
humsana-local
# If that doesn't work (PATH issues):
python -m humsana_local.server
⚙️ Add to Continue.dev
Add to ~/.continue/config.json under contextProviders:
{
"name": "http",
"params": {
"url": "http://localhost:25060/api/v5/context",
"title": "myself",
"displayTitle": "My State"
}
}
✨ Use It
Type @myself in Continue chat. Your AI now sees:
## My Current State
- State: deep_coding
- Focus Score: 0.87 (high - don't over-explain)
- Energy: 0.65 (good)
- Cognitive Load: deep_focus
- Interruptible: no - in the zone
AI Guidance: User is in deep focus. Be concise. Skip explanations. Provide code directly.
🔍 What It Detects
| Signal | How | Why It Matters |
|---|---|---|
| Focus Score | Window switching frequency | Lots of alt-tabbing = fragmented attention |
| Energy Level | Time of day + session length | 3am after 6 hours = probably exhausted |
| Coding Ratio | Time in IDE vs other apps | High ratio = deep in code |
| State | Combined signals | deep_coding, focused_work, context_switching, low_energy |
🔒 Privacy
- 100% local — nothing leaves your machine
- No accounts — just run it
- No telemetry — we don't even know you're using it
- Open source — read the code yourself
🍎 macOS Note
For window tracking, you may need to grant Terminal Accessibility permissions in System Preferences → Privacy & Security → Accessibility.
🆚 Local vs Cloud
| Feature | Humsana Local (Free) | Humsana Cloud |
|---|---|---|
| Privacy | 100% offline | Enterprise-grade encryption |
| Focus Detection | ✅ Window switching | ✅ + behavioral context |
| Energy Estimate | ✅ Time of day | ✅ + calendar + workload |
| Meeting Awareness | ❌ | ✅ |
| Communication Style | ❌ | ✅ |
| Personality Model | ❌ | ✅ OCEAN profiling |
| Burnout Prediction | ❌ | ✅ |
| Anomaly Detection | ❌ | ✅ |
Local is for Focus. Cloud is for Context.
Local sees you typing fast. Cloud knows it's the deadline, not the coffee.
💬 Feedback?
Early beta. Found a bug? Have an idea?
Open an Issue or DM on Twitter/X
License
MIT
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file humsana_local-1.0.0.tar.gz.
File metadata
- Download URL: humsana_local-1.0.0.tar.gz
- Upload date:
- Size: 6.3 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.11.1
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
d436fcdd200b40aebfbd172f1f8ea3e8cf222500106cfb56e78bfc67f86825e1
|
|
| MD5 |
42a5d51bd65d687c73339389168cb46b
|
|
| BLAKE2b-256 |
e2cc1959115881580793afd2da2f50cb8866d0604d56dd38680947e2d112cf27
|
File details
Details for the file humsana_local-1.0.0-py3-none-any.whl.
File metadata
- Download URL: humsana_local-1.0.0-py3-none-any.whl
- Upload date:
- Size: 6.9 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.11.1
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
b8797c958aa3ab3766a4f3e82a9dad94ca8417fe7aee150c3e7b4ddc7c189353
|
|
| MD5 |
3e00a5273b3376b7850e957a83a81a58
|
|
| BLAKE2b-256 |
1612375341972c086cc0973ccd8dcb94f58e94be1d2a971c94e948ccc2d680db
|