Skip to content

Jev_steer_or_queue

A Claude Code plugin, with a separate Codex folder, that classifies each message you send while the agent is working as steer, queue or interrupt, and holds the queued ones in a per-session FIFO for replay after the turn.

Stars
15
Language
Python
License
MIT
Published by
Community
Author
Larkspur-Wang
Last verified
Oct 2026

About this project

Your coding agent is busy. You send another message. Should it change course, wait its turn, or stop? A hook that lets TypeSafe Jev make that call in about half a second.

Quoted from the project's README. Read the full README.

Highlights

  • Queue and interrupt only fire at probability 0.9 or above, and interrupt also needs an explicit stop signal; anything else keeps native behaviour.
  • Messages sent while the agent is idle are never intercepted and never sent to Jev, so the common case costs nothing.
  • It fails open: a missing key, a network error or the two-second timeout means the message passes through untouched.
  • Shadow mode is the default, so decisions are logged for a few days before anything is actually held or stopped.

Watch out

The hook can only stop at a tool boundary, so a running command finishes first and a turn that is only producing text cannot be stopped at all. Pressing Esc fires no hook, so anything still queued is listed and dropped on your next prompt, and held messages never appear as user messages, only as the hook's notice and Stop hook feedback. Codex can be nudged but not blocked. For each mid-turn message it sends TypeSafe that message up to 2,000 characters and the start of the current task up to 500, nothing else. The README names colloquial Chinese, negations and multi-intent messages as the weak spots and tells you to calibrate thresholds on your own wording. Copy reviewed against the repository README on 2026-10-04, not from running it.

Questions about this entry

Short answers drawn from the linked project or page.

More Jev resources like this

51GitHub stars
Local LinkedIn recruiting workspace on Browser Use's Jev Ultrafast: Jev screens titles, drives navigation, and assesses profile excerpts against your brief while a live browser stays visible; matches and evidence export as JSON.
Jev Agents & MCP#community#jevbest#catalog-import
33GitHub stars
Codex Skill plus stdlib-only Python CLI that recommends which installed skill to read next for a task, allowing no-skill or uncertain answers and never executing candidates.
Jev Agents & MCP#community#jevbest#catalog-import
72GitHub stars
MCP server over stdio that gives coding agents fourteen Jev judgment tools, turning evidence plus an enumerable question into auto, review, or escalate actions from calibrated probabilities.
Jev Agents & MCP#community#jevbest#catalog-import
Research runtime that pages a large logical tool and context space through a small bounded Jev decision surface with explicit approvals.
Jev Agents & MCP#community#jevbest#catalog-import
25GitHub stars
Agent-oriented playbook for using Jev as a cheap, fast decision layer that triages, classifies, scores, and gates work before expensive browser, research, retry, or subagent steps.
Jev Agents & MCP#community#awesomejev#catalog-import
14GitHub stars
Opt-in Hermes plugin that routes each turn to one skill: a wide Jev Choice plus request checks gate a narrow fits round, injecting a single ignorable hint line.
Jev Agents & MCP#community#jevbest#catalog-import
Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai