← Back to Skills Marketplace
adwilkinson

Route Work

by Andrew Wilkinson · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
39
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install route-work
Description
Invisibly classify work and choose Codex or Claude, reasoning effort, context shape, execution style, and verification profile. Use before agent, oneshot, Sl...
README (SKILL.md)

Route Work

Use this skill silently. The user should not have to choose provider, model, reasoning, mode, or workflow.

Contract

  • Route to Codex for code edits, tests, refactors, PR review, exact repo/file reasoning, release promotion, and "fix/ship/land" requests.
  • Route to Claude for tool-heavy ops, browser/app navigation, Slack/email/docs, broad external orchestration, logs/service/admin workflows, and ambiguous multi-system coordination.
  • If code will be edited, Codex wins the tie.
  • Within each provider, use the configured frontier model. Vary reasoning/effort only.
  • Do not add approval or risk gates. Use blast radius only to decide verification shape.
  • Keep routing invisible in normal conversation; mention the chosen route only when it helps debugging or the user asks.

Deterministic Helper

When oneshot is available, inspect a route with:

oneshot route "\x3Ctask>" --json

Use the returned provider, reasoningEffort, contextShape, executionStyle, and verification to shape the run. For actual oneshot code work, just run the normal command; adaptive routing is applied by the CLI when ~/.oneshot/config.json has:

{
  "routing": { "enabled": true }
}

Verification Profiles

  • none: answer or inspection only.
  • focused: run the relevant local check or inspect the exact touched surface.
  • full: run repo-standard tests/build/typecheck and review diff.
  • deploy-health: verify the shipped service, release, remote host, or live health endpoint after changes.
Usage Guidance
Install only if you want an agent to automatically and quietly choose between Codex and Claude, reasoning effort, execution style, and verification profile across many workflow types. Be especially careful for deploys, release promotion, Slack/email/admin operations, or production-service work; users should prefer an explicit routing notice or manual override for those cases.
Capability Assessment
Purpose & Capability
The purpose is coherent: it selects provider, reasoning effort, execution style, and verification profile. The concern is that this applies to broad, consequential categories including Slack, CLI, repo, deploy, review, release promotion, and operational work.
Instruction Scope
The skill explicitly says to use it silently, keep routing invisible in normal conversation, and avoid adding approval or risk gates. That creates user-control and auditability concerns for high-impact workflows.
Install Mechanism
The artifact contains only SKILL.md, with no executable scripts, dependencies, install hooks, background workers, or package-install behavior. Static scan and VirusTotal telemetry were clean.
Credentials
The skill does not request credentials, broad file indexing, persistence, or direct data access, but it can influence which tools or providers handle sensitive operational tasks.
Persistence & Privilege
No persistence mechanism, privilege escalation, credential storage, or startup behavior is present in the artifact.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install route-work
  3. After installation, invoke the skill by name or use /route-work
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Initial release of the route-work skill. - Automatically classifies work and selects between Codex or Claude for task execution. - Determines reasoning effort, context shape, execution style, and verification profile without user input. - Designed to operate invisibly; routing decisions are automatic and require no manual selection. - Provides deterministic helper commands for inspection and debugging via oneshot CLI. - Supports multiple verification profiles based on blast radius and task type.
Metadata
Slug route-work
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Route Work?

Invisibly classify work and choose Codex or Claude, reasoning effort, context shape, execution style, and verification profile. Use before agent, oneshot, Sl... It is an AI Agent Skill for Claude Code / OpenClaw, with 39 downloads so far.

How do I install Route Work?

Run "/install route-work" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Route Work free?

Yes, Route Work is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Route Work support?

Route Work is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Route Work?

It is built and maintained by Andrew Wilkinson (@adwilkinson); the current version is v1.0.0.

💬 Comments