hermes-jev-plugin
Hermes Agent plugin exposing Jev as four coding-workflow tools: jev_check, jev_route, jev_score and jev_evaluate.
Overview
This plugin exposes TypeSafe's Jev to Hermes Agent as four tools focused on coding-agent workflows. jev_check asks atomic yes/no questions (Noul), jev_route picks one option (Choice) such as which subsystem owns a bug or which agent handles a task, jev_score rates on an ordered rubric such as severity, complexity or risk, and jev_evaluate asks any mix of questions about the same state in one cheap parallel call.
A bundled skill covers atomic question design, state preparation and confidence gating. The README includes recipes, such as failure triage where a raw error is the state and Nouls ask whether the failure is transient or a missing dependency (retry, fix versions or escalate), and a review gate over a diff asking whether a senior engineer would block it. Settings cover the model, HTTP timeout and maximum questions per evaluate call.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Is this failure transient (retrying could succeed)? | yes · no |
| Yes / no | Is this a missing dependency or version conflict? | yes · no |
| Yes / no | Would a senior engineer block this diff as-is? | yes · no |
| Choice | Which subsystem owns this bug / which agent handles this task? | — |
How it works
- Install the plugin and set TYPESAFE_API_KEY
- Agent calls jev_check, jev_route, jev_score or jev_evaluate
- Jev returns typed answers with confidence
- Agent branches (retry, fix, escalate, block) on the answers
Strengths
- Simple tool surface for common coding decisions
- Bundled question-design skill
- Batching via jev_evaluate
Caveats
- Agent must choose to call the tools
- Requires a TypeSafe key
Getting started
hermes plugins install ajensenwaud/hermes-jev-plugin (it prompts for TYPESAFE_API_KEY).At a glance
- Type
- Plugin
- Status
- beta
- Language
- Python
- Runtime
- Hermes Agent plugin
- Has code
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
TypeSafe Jev (System One) decision tools for Hermes Agent: jev\_check / jev\_route / jev\_score / jev\_evaluate
logicrw/awesome-jev-projectsTypeSafe Jev (System One) decision tools for Hermes Agent: jev_check /…
wh000wh000/awesome-jev-liveTypeSafe Jev (System One) decision tools for Hermes Agent: jev_check / jev_route / jev_score / jev_evaluate.
hellogumbo/awesome-jev
Frequently asked questions
Which tools are added?
jev_check, jev_route, jev_score and jev_evaluate.
Is there guidance on writing questions?
Yes, the bundled jev:jev-questions skill.
How many questions per call?
Up to 20 by default in jev_evaluate.
Related projects
CLI, MCP server and Claude Code plugin for closed-form yes/no, choice, score and rank decisions via Jev, OpenAI Decisions or Claude.
MCP serverbetaTypeScriptGives Claude Code, Codex, Pi and OpenCode five typed Jev tools (classify, check, score, rank, ask) plus a skill on when to use them.
MCP serverbetaTypeScriptMCP server exposing Jev Noul, Choice and Score via OpenRouter so coding agents can make quick decisions without reasoning tokens.
MCP serverbetaTypeScriptJava/Spring Boot MCP server exposing Jev classify, score, check and patch-gate tools over HTTP instead of stdio.
MCP serverbetaJavaClaude Code skill that hands bulk 'read a lot to decide a little' judgments over files, logs and tickets to Jev and returns a shortlist.
Agent skillbetaJavaScriptInstallable agent skills for triage, routing, code review and document work with Jev, plus 108 scenarios and recorded examples.
ArticlebetaPython