DataJev
Data-analysis agent where an LLM runs Python analysis and Jev decides whether to continue, switch direction, verify or stop.
This summary is based on how curated lists describe the project; its README was not available to review.
Overview
DataJev pairs a System 2 LLM data-analysis agent with Jev as System 1 control. According to the list summaries, the LLM performs Python-based analysis while Jev reads a compressed summary of the analytical state and decides the agent's next move: continue the current direction, switch to another one, verify a finding, or stop and synthesise the answer.
The idea is to keep the expensive model focused on analysis while a cheap, fast classifier handles loop control, a frequent source of wasted steps in agentic data work. No README was available, so implementation details are not known.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Choice | Given the current analysis state, what should the agent do next? | continue · switch · verify · stop |
How it works
- LLM runs a Python analysis step
- State is compressed into a summary
- Jev chooses continue, switch, verify or stop
- Loop ends with a synthesised answer when Jev says stop
Strengths
- Cheap loop control separate from the analysis model
- Keeps the expensive analysis model focused on analysis
Caveats
- List-only; no README reviewed
- Unknown evaluation
At a glance
- Type
- Project
- Status
- prototype
- Language
- Python
- License
- MIT
Use cases and integrations
How curated lists describe it
Data analysis agents: an LLM performs Python-based analysis while Jev reads the compressed analytical state and decides whether the agent should continue the current direction, switch to another one, verify a finding, or stop and synthesize the answer.
yibie/awesome-jev⚡ DataJev LLM → Analyze Jev → Continue / Switch / Verify / Stop System-1 control for System-2 data agents
logicrw/awesome-jev-projects
Frequently asked questions
What does Jev decide?
Whether the data agent should continue, switch direction, verify a finding, or stop.
Who does the analysis?
An LLM writing and running Python.
Why split the roles?
Loop control is a small typed decision, so a fast classifier can make it cheaply.
Related projects
AutoGPT agent platform, listed as adding Jev decision blocks for typed routing, filtering, scoring and confidence-gated next actions.
ProjectproductionPythonMulti-agent orchestration layer for Claude Code with opt-in Jev hooks for fast judgment points, decision caching and egress controls.
PluginproductionTypeScriptOh My Pi coding agent (a Pi fork with IDE wiring) with an optional TypeSafe judgment provider for bounded decisions.
ProjectproductionRust- SystemOneHarness★ 193
Agent harness that compiles an environment's finite actions into typed questions for Jev and gates each step by confidence.
CLIbetaPython Agent plugin that asks Jev whether the session is at a safe boundary to /compact and can trigger it on Pi and Claude Code.
PluginbetaTypeScriptChrome extension port of Jev Ultrafast that drives your current tab, with Jev picking the operation and element each step.
Browser extensionbetaTypeScript