User repository on GitHub
A local-first brain-budget meter for Codex: meme moods, local Laya scoring
- Forks
- 5
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Tools & CLIs
Headroom
A local-first Codex plugin and skill that gives AI conversations a daily brain budget, scored locally by Laya rather than Jev. Windows tray card, optional macOS menu bar or web dashboard, mood memes, and English and Chinese interfaces; Python 3.10+.
- Stars
- 5
- Language
- Python
- License
- MIT
- Reading time
- 3 min
- Published by
- Community
- Author
- llm-learner
- Last verified
- Oct 2026
Works forJev for Scoring
Used inJev for Software & DevTools
Highlights
- Daily limit = the busiest day in the previous seven complete days across supported tools, times two; the card shows percent left, reset at UTC+8.
- Codex scores automatically once you trust its SessionStart and UserPromptSubmit hooks; other supported tools contribute local history statistics only.
- Without hooks, python skills/headroom/scripts/headroom.py status estimates usage as today's messages times two; hook scores range from 0 to 10.
- Local by default, with English and Chinese interfaces; viewing, refreshing and clip playback never spend points and 0 percent never blocks conversations.
- Requires Python 3.10+ and readable local AI conversation history; the animated Windows tray card needs the Edge WebView2 Runtime.
Quickstart
bash
sh skills/headroom/hooks/install.sh --link-skill
python3 skills/headroom/scripts/headroom_dashboard.py --lang enWatch out
Copy reviewed against the repo README on 2026-10-15. The README itself labels the meter for fun, not science - it measures neither intelligence, health nor fatigue, automated-task filtering is best-effort, and no accuracy or performance benchmarks are published.
Questions about this entry
Short answers drawn from the linked project or page.
A daily brain budget you define for yourself, not an AI usage limit. It pools the busiest day in the previous seven complete days across supported tools and multiplies by two; the display shows today's percent left and days reset at UTC+8. The README stresses it is for fun, not science.
No key is needed. It is local-first: it reads readable local AI conversation history on your machine and defaults to local scoring. Viewing, refreshing and clip playback never spend points, and the README says the meter is not a measurement of intelligence, health or fatigue.
Clone the repo, then run sh skills/headroom/hooks/install.sh --link-skill and start the dashboard with python3 skills/headroom/scripts/headroom_dashboard.py --lang en. Open http://127.0.0.1:8766/?lang=en. macOS additionally has an optional menu bar display documented in the guide.
Installing the plugin or skill alone does not enable hooks. Open /hooks in Codex, review and trust headroom's SessionStart and UserPromptSubmit hooks, then start one new session, keeping the repository in place. On Windows the installer stops rather than overwrite an existing hooks.json.
Codex gets automatic scoring after you trust its hooks. Claude Code, opencode, Antigravity and WorkBuddy contribute local history statistics but are not configured automatically; their usage is estimated as today's messages times two unless you configure their hooks separately.
No. The README states zero percent never blocks your conversations; the meter is a nudge, not a gate. Without hooks you can still check status with python skills/headroom/scripts/headroom.py status, which estimates usage from message counts.
llm-learner built it as a Codex plugin and skill; the madewithjev listing describes it as scored by Laya, the open System 1 decision models, with hook scores ranging from 0 to 10. It ships English and Chinese interfaces, and the animated Windows tray card needs the Edge WebView2 Runtime.
More Jev resources like this
An Ollama-style local runner for System One decision models from Hugging Face: pull, serve and query them behind the same /v1/systemone API as TypeSafe's Jev, with a terminal model manager and a browser playground.
Jev Tools & CLIs#community#catalog-import#madewithjev
Community
nvkudva/ollajevPython
nvkudva/ollajevPythonA self-hosted System One service that runs Laya through MLX on Apple silicon and keeps the model's native request and response format, exposing fast typed decisions to your own devices over a private Tailscale HTTPS endpoint.
Jev Tools & CLIs#community#github#macos
Community
arhamj/system-onePython
arhamj/system-onePythonA local Telegram conditional filter: FastAPI and SQLite listen on 127.0.0.1:18721, tasks batch one account's messages to TypeSafe Jev or a local Laya model, and only hits are shown, with a UI in Traditional Chinese, Simplified Chinese, and English.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
wing9897/jev.tgPython
wing9897/jev.tgPythonAgent tooling that turns a model idea into a buildable LEGO model: the agent writes a plan and a generator script that produce LDraw source, plus a 3D viewer, a Blender-editable glTF file and a Meta Quest 3 VR view. Part search uses jev-rerank, a semantic search reranked by TypeSafe's Jev.
Jev Tools & CLIs#community#catalog-import#madewithjev
Community
anteloc/ldraw-novaPython
anteloc/ldraw-novaPythonA desktop app for dictation, audio-file transcription and meeting notes on Windows, macOS and Linux, using local Whisper models or the OpenAI API, with an opt-in MCP server and local history API so agents can search saved transcripts.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
knuckles92/openwhisperPython
knuckles92/openwhisperPythonA Mac-only voice assistant: say Hey Jev or hold right Option and Jev decides every command in one call for $0.00004 per request, while Whisper transcribes locally, Fish Audio S2.1 Pro speaks replies, and an LLM wakes only for questions.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
henryklunaris/hey-jevPython
henryklunaris/hey-jevPythonFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.