Organization repository on GitHub
Jev based decision tools for claude code
- Forks
- 5
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Tools & CLIs
Optim Jev
A Claude Code plugin from Oqura-ai that spends fewer tokens in long sessions with three Jev-powered tools: jev-compacter removes stale tool calls and outputs, jev-router picks a model per file change, and jev-skills sends a short skill list per prompt, each with a stats command.
- Stars
- 33
- Language
- Python
- Reading time
- 2 min
- Published by
- Community
- Author
- Oqura-ai
- Last verified
- Oct 2026
Used inJev for Software & DevTools
- Languages
- Python
- Capabilities
- Open sourceRequires API key
Highlights
- Three tools in one plugin: jev-compacter is on by default, while jev-router and jev-skills stay off until enabled.
- jev-compacter removes stale tool calls and outputs instead of summarizing the whole conversation away.
- jev-router routes file changes to the right model using project-specific rules that evolve with your workflow.
- jev-skills sends a short, relevant skill list for each prompt instead of the entire catalog.
- Each tool works independently, fails open, and has a stats command reporting session-level cost and savings in US dollars; badges say Python 3.10+.
Watch out
README and GitHub metadata reviewed 2026-10-17; the digest truncates the jev-compacter detail block, so setup, configuration, and measured savings were not available. The README badge says MIT but the GitHub license field in the digest is empty, so check the repository before reusing the code.
Questions about this entry
Short answers drawn from the linked project or page.
A Claude Code plugin from Oqura-ai that combines Jev-powered context compaction, model routing, and skill selection to reduce token spend in long sessions without changing how you work.
All three tools are described as Jev-powered: jev-compacter asks Jev what stale tool output can go, and jev-router and jev-skills use Jev decisions to route file changes and select skills. The capture does not show prompts or endpoints.
Only jev-compacter is on out of the box; jev-router and jev-skills ship off. Each tool works independently, so you can enable them one at a time.
The README says each tool includes a stats command that reports session-level cost and savings in US dollars, so you can see whether the added decisions pay for themselves.
Each tool fails open, per the README, meaning your Claude Code session keeps working without the optimization rather than breaking.
The badges list Python 3.10 or newer, and because the tools are Jev-powered they call TypeSafe's hosted model, so a Jev key and account are implied; the capture does not show the quick-start steps.
The README carries an MIT badge, but the GitHub license field in the digest is empty, so the capture cannot confirm the terms; check the repository before redistributing or reusing the code.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
YouTube: Ray Amjad puts Jev inside an agentic coding loop and looks at what it costs to run.
YouTube: RoboNuggets puts Jev to work inside Claude Code.
Reddit: A Claude Code mod replaces the auto-mode classifier with Jev, 2x faster.
More Jev resources like this
Claude Code plugin whose UserPromptSubmit hook asks Jev a Choice over /effort levels (low, medium, high, max, unclear) plus a Noul on whether a hands-off request has a fuzzy spec, then offers a one-key switch before Claude starts when the tip clears its confidence threshold.
Jev Tools & CLIs#community#claude-code#plugin
A proof-of-concept Python scanner that maps a backend codebase's attack surface and ranks likely vulnerabilities down to the suspicious line, using Jev as the classifier for about one cent per repo. It reads Python, JavaScript and TypeScript, and a demo site shows read-only results on PyGoat and NodeGoat.
Jev Tools & CLIs#community#github#security
Gmail triage with TypeSafe Jev: every few minutes it reads new inbox mail, asks Jev for a category with calibrated confidence, then labels, archives, unsubscribes from verified marketing or cold outreach, and moves scams to Spam.
Jev Tools & CLIs#community#github#gmail
Community
hellolitebox/litebox-inbox-jevPython
hellolitebox/litebox-inbox-jevPythonAn MIT-licensed case study with two working Python programs pairing Jev and Claude Code: a support-triage app with Jev inside it, and a Claude Code safety hook where Jev judges the agent's own shell commands. Tests and evals run offline, free.
Jev Practices & Patterns#community#github#python
An independent open-source pipeline that gives Jev vision: it detects and tracks objects in images, video, and RGB-D camera feeds, then asks Jev a typed question about each relevant object and prints the chosen answer with its confidence.
Jev Tools & CLIs#community#github#python
Community
CharlesFeng0314/JEV_seesPython
CharlesFeng0314/JEV_seesPythonA proof-of-concept HTTP service for constrained decisions with a local MLX model: send text or JSON plus typed choice, score, or noul questions and it returns probabilities over the answers you declared — no generated text, no parsed model JSON.
Jev Tools & CLIs#community#github#python
Community
snellingio/system-onePython
snellingio/system-onePythonFrom 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.