Get the FREE Ultimate OpenClaw Setup Guide →

router-first-architecture

npx machina-cli add skill parcadei/Continuous-Claude-v3/router-first-architecture --openclaw
Files (1)
SKILL.md
1.6 KB

Router-First Architecture

Route through domain routers before using individual tools. Routers abstract tool selection.

Pattern

Domain routers (like math-router) provide deterministic mapping from user intent to exact CLI commands. Always use the router first; only bypass for edge cases.

DO

  • Call math-router route "<intent>" before any math operation
  • Let domain skills co-activate with their router (via coActivate in skill-rules.json)
  • Trust the router's confidence score; only fall back if command: null
  • Keep trigger keywords/patterns in skill-rules.json broader than routing patterns

DON'T

  • Call individual scripts directly when a router exists
  • Duplicate routing logic in individual skills
  • Let domain skills bypass their router

Co-Activation Pattern

Domain skills should co-activate with their router:

{
  "math/abstract-algebra/groups": {
    "coActivate": ["math-router"],
    "coActivateMode": "always"
  }
}

This ensures the router is always available when domain knowledge is activated.

Two-Layer Architecture

  1. Skill-rules trigger layer: Nudges Claude to use the router (keywords, intent patterns)
  2. Router routing layer: Deterministic mapping to scripts via regex patterns

Keep the trigger layer broader than routing - the router should handle "not found" gracefully.

Source Sessions

  • 2bbc8d6e: "Trigger layer was narrower than routing layer" - expanded triggers
  • This session: Wired 8 domain math skills to co-activate with math-router

Source

git clone https://github.com/parcadei/Continuous-Claude-v3/blob/main/.claude/skills/router-first-architecture/SKILL.mdView on GitHub

Overview

Router-First Architecture routes user intent through domain routers before invoking individual tools. Routers provide deterministic mappings from intent to exact CLI commands, ensuring consistent, explainable tool selection. The pattern uses a two-layer setup: a trigger layer to nudge routing and a router layer to perform the mapping, with co-activation to keep routers available.

How This Skill Works

Before any math operation, call the domain router (e.g., math-router) with the intent via route. The router returns a command and a confidence score; if a command is provided, execute it. If command is null, fall back to edge-case handling. Domain skills co-activate with their router (via coActivate in skill-rules.json), ensuring the router is always available.

When to Use It

  • You have a clear intent that maps to a specific CLI command via a domain router (e.g., math-router).
  • Starting a math operation in a session where multiple tools could be applicable and you want deterministic routing.
  • Coordinating routing across several domain skills in a multi-skill workflow.
  • Avoiding duplication of routing logic by delegating to the router instead of in-skill routing.
  • When the router has a valid command returned (not null) and should govern execution; edge cases require bypass.

Quick Start

  1. Step 1: Call the router before any math operation using math-router route \"<intent>\"
  2. Step 2: Ensure domain skills co-activate with their router via coActivate in skill-rules.json
  3. Step 3: If a command is returned, execute it; if command is null, apply the edge-case fallback

Best Practices

  • Always call the router first with the intent: e.g., math-router route \"<intent>\" before any tool invocation.
  • Keep trigger keywords broader than routing patterns to improve router coverage.
  • Enable co-activation of domain skills with their router using the coActivate pattern so the router is always available.
  • Trust the router's confidence score and only bypass when command is null.
  • Avoid duplicating routing logic in individual skills; let routers handle the mapping.

Example Use Cases

  • A user requests to simplify a polynomial; the router maps the intent to the exact CLI command via math-router and returns a concrete script to run.
  • In a session with abstract algebra tasks, the router directs to the appropriate domain tool (e.g., mapping 'groups' related intents to a specific script).
  • Edge-case handling where the router cannot map the intent (command: null) triggers a fallback path to a manual script or guidance.
  • Two-domain coordination where math-router and another domain router co-activate to ensure routing is available during multi-skill tasks.
  • A sequence like 'solve differential equation' is routed to the precise command by the router, then executed deterministically.

Frequently Asked Questions

Add this skill to your agents
Sponsor this space

Reach thousands of developers