Agent skill

Jevrouter

by BillionsBobby in BillionsBobby/JevRouter

Use Jev to choose among relevant tools, models, skills, MCP servers, and Codex subagents for non-trivial task steps when routing can save execution time or reasoning cost.

MITAuto-check passedAgent Workflows

Install Jevrouter

skills CLI
$ npx skills add BillionsBobby/JevRouter --skill jevrouter -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install BillionsBobby/JevRouter jevrouter --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/BillionsBobby/JevRouter.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/jevrouter .claude/skills/jevrouter && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
jevrouter
GitHub stars
450
Token cost
~839 tokens
SKILL.md length
389 words
Files
1
Skills in repo
1
Repo updated
First seen
Licence
MIT

At a glance

Use Jev to choose among relevant tools, models, skills, MCP servers, and Codex subagents for non-trivial task steps when routing can save execution time or reasoning cost.

  • Tasks that involve Subagents
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
  • Tasks that involve MCP servers

What it does

Jevrouter is an agent skill from BillionsBobby/JevRouter. Use Jev to choose among relevant tools, models, skills, MCP servers, and Codex subagents for non-trivial task steps when routing can save execution time or reasoning cost. Use the real current task and available candidates.

Its SKILL.md is about 840 tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Agent Workflows, covering Subagents and MCP servers. It works with Model Context Protocol. The repository describes itself as: A lightweight Jev-powered router for models, tools, and subagents. The licence is MIT.

When your agent uses it

  • Tasks that involve Subagents
  • Tasks that involve MCP servers

Example prompts

  • “/jevrouter”

What it can do on your machine

Read from SKILL.md and the folder at commit b1cdf80. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are bash).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Jevrouter loads about 839 tokens when it runs. Until then it costs about 58 tokens; SKILL.md has 389 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~58
When it runs · the whole SKILL.md, loaded when a task matches
~839

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from BillionsBobby/JevRouter at commit b1cdf80, republished under its MIT licence (© BillionsBobby). 389 words, ~839 tokens.

Download SKILL.mdSave it as .claude/skills/jevrouter/SKILL.md (or your agent's skills folder).
name
jevrouter
description
Use Jev to choose among relevant tools, models, skills, MCP servers, and Codex subagents for non-trivial task steps when routing can save execution time or reasoning cost. Use the real current task and available candidates.

JevRouter

Announce “JevRouter: routing the next step” when starting. A past setup check is not a routing decision for the current task.

Route each non-trivial step when there are materially different available ways to do it and Jev's selection can reduce tool choice or reasoning overhead. Include commands, MCP tools, skills, and Codex subagents when they are actually available and relevant. Keep the candidate set small and task-specific (usually 2–5), rather than sending the entire installed tool/agent catalog. Skip trivial chat, fixed one-operation requests, and cases where the routing call would cost more time than it can save. For Codex subagents, use exact name and description values from relevant ~/.codex/agents/*.toml profiles or built-in runtime agents; never send their developer_instructions to Jev. Jev selects; the Codex host performs the actual native spawn and enforces permissions.

Use the installed CLI helper. It reads the inherited Jev key. Do not print keys, put them in arguments, or replace them with placeholders.

sh
{{JEVROUTER_COMMAND}} route --stdin <<'JEV_INPUT'
{
  "request": "Find the latest interview and retrieve its original source",
  "context": {"next_step": "Find the source before summarizing it"},
  "candidates": [
    {"name": "web_search", "description": "Find current web sources"},
    {"name": "summarize", "description": "Summarize sources already retrieved"}
  ]
}
JEV_INPUT

The names above illustrate the shape: replace the task and candidates with this turn's real capabilities. Model/subagent candidates use {id,name,type,description} with type: model or subagent. To discover Codex profiles, use jevrouter discover --codex-agents "$HOME/.codex/agents" --codex-agent-names agent_one,agent_two and route only the task-relevant profile manifests. CLI also accepts --request with --candidates-file for JSON/YAML, and SDK users can call route({request,candidates}).

Read the JSON on stdout; progress is on stderr. State the returned decision_id, status, decision.selected, and whether the provider was live or demo. If status=selected, execute only that available capability through the host's own permission checks. If confirmation is required, obtain it first. On no_decision, invalid input, missing credentials, or network error, explain the fallback and gather better inputs; never describe a failed or demo call as live Jev routing. CLI exit codes: 0 selected, 2 review/no-decision, 1 error.

Show full SKILL.md (95 more words)Show less

After execution, include the observation in the next routing request when needed. This is a routing aid: it does not switch the host's underlying model, spawn agents, or execute returned IDs automatically. For a selected subagent, use the host's native spawn capability with the selected agent type and return its result to the parent task. Do not recurse into routing the router or run routing again just to make status green.

If the optional MCP tool jev_route is already loaded, it accepts the same request and candidate payload. Always respect project instructions and host authorization.

© BillionsBobby, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/jevrouter of BillionsBobby/JevRouter.

Open the folder on GitHubat commit b1cdf80

Compare with similar skills

Jevrouter next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Jevrouter compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Jevrouter this skillBillionsBobby/JevRouter450—~839Automated safety check: PassMIT
Claude Automation Recommenderanthropics/claude-plugins-official38k3 repos~2.7kAutomated safety check: NotesApache-2.0
CC Workflow Studio AI Editorbreaking-brake/cc-wf-studio5.4k—~561Automated safety check: PassCustom licence
Agent Deckasheshgoplani/agent-deck1k—~1.7kAutomated safety check: PassMIT
Bootstrapinfragate/capa724—~5.9kAutomated safety check: PassMIT
Puppetmaster Agent Orchestrationprofessorpalmer/Puppetmaster467—~3.2kAutomated safety check: PassMIT

Similar skills

  • Claude Automation Recommender

    anthropics/claude-plugins-official

    Official

    Scans a codebase and suggests which Claude Code hooks, subagents, skills, plugins and MCP servers fit its stack, without changing any files.

    38k GitHub starsUsed in 3 repos~2.7k tokens
    Agent WorkflowsAuto-check: notes
  • CC Workflow Studio AI Editor

    breaking-brake/cc-wf-studio

    Creates and edits visual agent workflows in CC Workflow Studio through conversation, with the agent reading and writing the canvas over MCP.

    5.4k GitHub stars~561 tokensUpdated 2 days ago
    Agent WorkflowsAuto-check passed
  • Agent Deck

    asheshgoplani/agent-deck

    agent-deck, the terminal session manager for AI coding agents.

    1k GitHub stars~1.7k tokensUpdated 3 days ago
    Agent WorkflowsAuto-check passed
  • Bootstrap

    infragate/capa

    Capify an existing project. An agent skill from infragate/capa.

    724 GitHub stars~5.9k tokensUpdated 6 days ago
    Agent WorkflowsAuto-check passed
  • Puppetmaster Agent Orchestration

    professorpalmer/Puppetmaster

    Operates and supervises Puppetmaster, a multi-agent orchestrator, through its MCP tools or CLI, picking the right verb for edits, reviews, audits and long-running jobs.

    467 GitHub stars~3.2k tokensUpdated today
    Agent WorkflowsAuto-check passed
  • Agentforce Generate

    SalesforceAIResearch/agentforce-adlc

    Build, modify, audit, repair, optimize, debug, and deploy agents with Agentforce Agent Script.

    114 GitHub starsUsed in 1 repo~7.4k tokens
    Agent WorkflowsAuto-check passed

Categories

Questions about Jevrouter

What does Jevrouter do?

Use Jev to choose among relevant tools, models, skills, MCP servers, and Codex subagents for non-trivial task steps when routing can save execution time or reasoning cost. Jevrouter is an agent skill from BillionsBobby/JevRouter. Use Jev to choose among relevant tools, models, skills, MCP servers, and Codex subagents for non-trivial task steps when routing can save execution time or reasoning cost.

When should I use Jevrouter?

Jevrouter fits situations like: tasks that involve Subagents; tasks that involve MCP servers.

How do I install Jevrouter in Claude Code?

Run `npx skills add BillionsBobby/JevRouter --skill jevrouter -a claude-code`. Or copy the skill folder (skills/jevrouter in BillionsBobby/JevRouter) into .claude/skills/jevrouter in your project. Claude Code loads it when a task matches its description.

How do I install Jevrouter in Codex?

Run `npx skills add BillionsBobby/JevRouter --skill jevrouter -a codex`. Or copy the skill folder (skills/jevrouter in BillionsBobby/JevRouter) into .agents/skills/jevrouter in your project. Codex loads it when a task matches its description.

Can I use Jevrouter in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add BillionsBobby/JevRouter --skill jevrouter -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/jevrouter, .gemini/skills/jevrouter, .github/skills/jevrouter and .opencode/skills/jevrouter in your project.

What does Jevrouter need to run?

SKILL.md names no scripts, command-line tools or credentials: Jevrouter is instructions for the agent only.

Does Jevrouter access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Jevrouter safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Jevrouter use?

Jevrouter is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Jevrouter use?

About 839 tokens (SKILL.md is roughly 3.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Jevrouter?

Skills that share tags, products or a category with Jevrouter: Claude Automation Recommender (anthropics/claude-plugins-official, 38k stars), CC Workflow Studio AI Editor (breaking-brake/cc-wf-studio, 5.4k stars), Agent Deck (asheshgoplani/agent-deck, 1k stars) and Bootstrap (infragate/capa, 724 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Jevrouter?

BillionsBobby (a GitHub user) maintains it in BillionsBobby/JevRouter, which has 450 GitHub stars. The repository was last updated on October 4, 2026.

Source: BillionsBobby/JevRouter on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.