GGhalebDweikatAgent tooling

winnow

A calibrated context sieve for Claude Code: every tool result is judged by a System One model before it enters context.

Explore repository

How it uses Jev

Terms used below. The judge is the model that answers one yes/no question per block ("is this block needed for the current task?") with a probability. By default that is Jev , TypeSafe AI's System One model: a model that returns calibrated probabilities for typed questions instead of generating text, so a hundred questions come back in one call in a few hundred milliseconds. Jev is in early access. The adapter is Ty…

Getting started

Requirements: Python 3.10+, uv, Claude Code 2.1.260 or newer with function hooks enabled (early access; one line in settings, below).

Source: community catalog. Details and metrics reflect the published community snapshot and are not a performance endorsement. Documentation details extracted from GitHub · the README.

Keep exploring

Similar to winnow

Browse all
dsh-jev-prune project preview

dsh-jev-prune

Agent tooling
Jev-judged context compaction for DeepSeek Harness: semantic tool-result pruning + deterministic receipt compaction
YMade byyangyu6663
fast-jev-compaction project preview

fast-jev-compaction

Agent tooling
Claude Code plugin that replaces the compaction summary with Jev decisions: every tool call and result is scored in one fast request, stale ones are dropped or truncated, everything kept stays verbatim.
TMade bytamaratran6.7k
jev-mcp (Songokou1983) project preview

jev-mcp (Songokou1983)

Agent tooling
Local MCP server exposing TypeSafe Jev (System One decision model) as native Claude Code / Codex tools
SMade bySongokou19832
pi-jev-context (Nyarlathoteppppp) project preview

pi-jev-context (Nyarlathoteppppp)

Benchmarks & research
Cache-neutral context trimming for the pi coding agent, powered by TypeSafe Jev: long tool output cut to verbatim key lines before it enters context, with lossless recall. Measured, with pre-registered benchmarks.
NMade byNyarlathoteppppp7
jev-filter project preview

jev-filter

Benchmarks & research
Semantic filtering for AI tool results. Native npm CLI, caller-controlled context, batching, and reproducible benchmarks.
AMade byapixly-ai1
open-jev (JoshuaSP) project preview

open-jev (JoshuaSP)

Jev alternatives & open source
Typed JSON inference with DiffusionGemma, with Every and Jev benchmark results
JMade byJoshuaSP39
Winnow project preview

Winnow

Browser & computer use
Know before you click. A Chrome extension that reads articles and YouTube videos ahead of you and says read, skim, save, or skip — with a confidence, tuned to your goals. Open source, MV3, powered by Jev.
TMade byThinkyMiner2
jev-agent-authorization project preview

jev-agent-authorization

Benchmarks & research
Jev agent authorization for MCP tool calls: Kinde identity and permissions plus Jev's typed, calibrated decisions, checked server-side before every call runs
KMade bykinde-starter-kits5
lejudge-jev-jepa project preview

lejudge-jev-jepa

Applications
Natural-language constraints for JEPA world-model planning, judged by a decision model instead of an LLM.
AMade byAbdelStark9