fast-dev-compaction

Codex plugin: verbatim Jev-guided context restoration around session compaction. Port of tamaratran/fast-jev-compaction to Codex lifecycle hooks.

How it uses Jev

Codex plugin that wraps session compaction with Jev decisions: before Codex compacts, every tool call and result is scored in fast requests; after compaction, the verbatim history Jev kept is re-injected as context. Also usable as an npm library.

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 fast-dev-compaction

Browse all
pi-fast-jev-compaction project preview
Pi extension: verbatim context compaction with TypeSafe Jev decisions
opencode-context-pruner project preview
Continuous verbatim context pruning for OpenCode, powered by TypeSafe Jev. Port of fast-jev-compaction adapted to OpenCode's context hook.
omp-jev-compaction project preview
Agent tooling
Verbatim Jev-scored context reduction for omp, over TypeSafe or OpenRouter
jev-dev project preview
Demos & playgrounds
同じ発言を jev と LLM の両方に判定させ、感情の変動値のズレと応答速度を1画面で見比べるデモ(affectus + Vercel AI Gateway)
laya-server (1Panel-dev) project preview
A self-hosted API and web interface for Laya’s structured decision models, compatible with the TypeSafe Jev API format.
jev-acp project preview
Integrations
Use Jev's Choice, Score, and Noul decisions from ACP-compatible clients, with guided input, reusable templates, and probability displays.
jevalyzer project preview
Benchmarks & research
Grade the agent sessions already on your disk. Claude Code, Codex, opencode, Gemini CLI and Antigravity, scored with Jev for cents.
typesafe-ai-playground (markjaquith) project preview
A playground for experiments around Jev, TypeSafe's System One model.
pi-jev-context (Nyarlathoteppppp) project preview
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.