Codex-Jev
Codex Decision: VS Code extension that trims repetitive lines from Codex tool output via Jev or OpenAI Decisions, linking the original.
Overview
Codex Decision selects repetitive lines for removal from supported local tool output in Codex. Errors and useful evidence stay visible, and every shortened result links to the complete original, so the agent spends less context on noise without losing access to anything.
Version 0.11.8 installs as a single VSIX in Windows VS Code with a Linux x86_64 WSL environment. It contains the hook, skill, session controls, settings and a verified Codex UI patch; on activation it installs or updates the local plugin and repairs supported Codex builds, with commands to check and repair the integration. You connect either OpenAI Decisions or TypeSafe Jev, entering and testing the key in a dialog so it stays out of chat. Standalone CLI setup is documented.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Is this line of tool output repetitive and safe to remove?(inferred) | yes · no |
How it works
- Codex runs a tool and produces output
- Decision provider selects repetitive lines
- Shortened output shown with a link to the full original
- Errors and evidence stay visible
Strengths
- Nothing lost: originals linked
- Provider choice (Jev or OpenAI Decisions)
- Self-repairing integration
Caveats
- Windows VS Code with WSL
- Patches the Codex UI
Getting started
code --install-extension codex-decision-0.11.8.vsix --force, then connect a provider in Connect Decision.At a glance
- Type
- Plugin
- Status
- beta
- Runtime
- VS Code extension + Codex hook (WSL)
- Has code
- yes
Use cases and integrations
How curated lists describe it
VS Code Codex Plugin for Jev-gated tool output integration.
hellogumbo/awesome-jev
Frequently asked questions
What does it remove?
Repetitive lines from tool output, keeping errors and evidence.
Can I see the full output?
Yes, every shortened result links to the original.
Which providers?
TypeSafe Jev or OpenAI Decisions.
Related projects
Codex hook that compacts large command and tool outputs, using Jev to rank which chunks matter and keeping the full output retrievable.
ServicebetaPythonRecursive language model layer for coding agents that keeps inputs in a local evaluator, with optional Jev judge_many for typed calls.
ProjectbetaRustCodex plugin that builds handoff packets by asking Jev whether each old read-only tool record should stay, keeping retained text verbatim.
ProjectbetaJavaScriptAgent plugin that asks Jev whether the session is at a safe boundary to /compact and can trigger it on Pi and Claude Code.
PluginbetaTypeScriptLocal compaction proxy for Claude Code and Codex that compacts more often with local rules and keeps full history searchable.
ProxybetaGoClaude Code (and Codex) plugin that trims long Bash output chunk by chunk with Jev before it reaches the model, without summarising.
PluginbetaTypeScript