omo-jev-plugin
OmO/senpi agent plugin where Jev or Span-01 rates which skills and next tools fit the task and spots loops or completion.
Overview
omo-jev-plugin connects structured decisions from Jev, or OpenRouter's Span-01 and Span-01 Lite, to OmO and senpi agents. It evaluates which skills and next tools fit the current task, looks for repetition or possible completion, and passes short suggestions back to the agent. It does not execute tools or replace senpi's permission checks.
It installs as an npm package through omo install, checks for updates (auto-update is opt-in), and creates a default config in advise mode. Keys come from the config file or TYPESAFE_API_KEY / OPENROUTER_API_KEY; without a key it warns and makes no requests. The README discloses telemetry: a pseudonymous install ID, session ID, provider and model, usage counts, and by default detailed per-decision counts, tokens and estimated cost, which can be reduced in config. A Korean README is available.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Score | How well does this skill or next tool fit the current work? | — |
| Yes / no | Is the agent repeating itself? | yes · no |
| Yes / no | Does the task look complete? | yes · no |
How it works
- Install with omo install npm:omo-jev-plugin.
- Set a Jev or Span-01 key in ~/.omo/jev-plugin.jsonc.
- In advise mode, the plugin scores skills, tools, repetition and completion.
- Short suggestions are passed to the agent, which decides.
Strengths
- Advisory only; no tool execution or permission bypass.
- Choice of Jev or OpenRouter Span-01 providers.
- Fails safe without a key.
Caveats
- Detailed telemetry is on by default.
- Suggestions may be ignored or misjudged by the agent.
Getting started
omo install npm:omo-jev-plugin, then add your API key to ~/.omo/jev-plugin.jsonc.At a glance
- Type
- Agent skill
- Status
- beta
- Language
- TypeScript
- Runtime
- OmO / senpi plugin
- Has code
- yes
Use cases and integrations
How curated lists describe it
Jev-powered decision support for OmO and senpi agents.
hellogumbo/awesome-jevRates how well the next skill or tool fits the work an OmO or senpi agent is doing and passes back a short suggestion. It runs no tool itself and replaces no permission check. Korean.
MrJev/awesome-jev
Frequently asked questions
Does the plugin run tools?
No, it only gives suggestions and does not replace permission checks.
Which providers are supported?
TypeSafe Jev or OpenRouter's Span-01 and Span-01 Lite.
Does it collect telemetry?
Yes; detailed telemetry is on by default and can be turned off in config.
Related projects
Claude Code function hook that sends the agent's this-or-that questions to Jev and optionally ranks installed skills per prompt.
ProjectbetaTypeScriptDeepSeek Harness plugin that registers a native jev_decide tool and declares its own policy for when the agent should ask Jev.
PluginbetaJavaScriptPortable decision layer for agent harnesses: Jev picks from host-supplied capabilities while the host keeps permissions and execution.
ProjectbetaTypeScriptLocal MCP server putting Jev in the coding loop: route next steps, pick prepared tool calls, review patches and verify claims.
MCP serverbetaTypeScript- SystemOneHarness★ 193
Agent harness that compiles an environment's finite actions into typed questions for Jev and gates each step by confidence.
CLIbetaPython - SkillRanker★ 134
Rust CLI that ranks which agent skills fit the next step from live session context with Jev, can abstain, and hooks into Claude Code.
CLIbetaRust