Jev Review
A code-review pipeline assembled from small typed judgments instead of one long reviewer prompt.
What Jev decides
What is risky, which files and evidence matter, what could fail, how severe it is, and who should review it.
Jev Review chains risk checks, file selection, evidence selection, mechanism classification, severity scoring, and reviewer routing, then presents inspectable findings in a local dashboard.
- Jev role
- Filter review scope, select evidence, score severity, and route candidate findings.
- Input
- Repository metadata, diffs or files, candidate snippets, and earlier pipeline decisions.
- Output
- A sequence of Noul probabilities, Choice distributions, and Score results.
- Decision frequency
- Across multiple stages of one repository review run.
- Fallback
- Human verification plus normal tests, linters, type checks, and security tools.
How it works
Reported performance
Overview
Jev Review decomposes repository review into a sequence of Noul, Choice, and Score calls. It first builds a risk matrix, narrows the files and snippets worth inspecting, asks for a failure mechanism and severity, and routes the finding to a suitable reviewer persona.
The README describes the tool as experimental. Findings are leads for a human reviewer, not proof of a bug, and the pipeline does not replace compilers, tests, static analysis, or security scanners.
- Language
- TypeScript
- License
- MIT
- Author
- Dev Agrawal
Sources
Related projects
Distill
Route coding-agent work, reasoning effort, and context retention through bounded Jev decisions.
What Jev decides
Which configured model and effort should handle a call, what evidence deserves attention, whether another attempt is useful, and what context should be retained.
jev-model-router
Route Claude Code model tiers and reasoning effort with one parallel Jev classification.
What Jev decides
Which fast, balanced, or deep tier fits the work, how much reasoning it needs, and whether production risk requires escalation.
jev-ultrafast
A browser agent where Jev selects an operation and a compatible page element from a live indexed action space.
What Jev decides
Which supported browser operation to run, and which compatible visible element to target