Skill

omni-routes

Manages agent routing rules: create per-chat or per-user routes to providers/agents, test route resolution, update/delete routes, and view metrics via omni CLI.

Bash

cli-tools

automation

npx claudepluginhub automagik-dev/omni --plugin omni

Popularity

Parent stars

Parent forks

Invocation

How this skill is triggered — by the user, by Claude, or both

Slash command

/omni:omni-routes

User invocable

Model invocable

Inline context

Default effort

Tool Access

This skill is limited to the following tools:

Bash(omni *)Bash(jq *)

Context Preview

The summary Claude sees in its skill listing — used to decide when to auto-load this skill

```bash

SKILL.md

84 lines · ~668 tokens

Similar Skills

omni-agent-setup

Guides connecting AI agents to Omni channels like WhatsApp, Telegram, Discord, Slack using providers, reply filters, routing, and testing via bash CLI.

2 tools

omni

openrouter-routing-rules

1.9k

Implements Python rules engine for OpenRouter model selection using user tier, task type, budget, tools, vision, and latency conditions with priorities and fallbacks.

7 files5 tools

openrouter-pack

model-routing-patterns

Model routing configuration templates and strategies for cost optimization, speed optimization, quality optimization, and intelligent fallback chains. Use when building AI applications with OpenRouter, implementing model routing strategies, optimizing API costs, setting up fallback chains, implementing quality-based routing, or when user mentions model routing, cost optimization, fallback strategies, model selection, intelligent routing, or dynamic model switching.

16 files5 tools

openrouter

Stats

Parent stars14

Parent forks1

MaintenanceExcellent

Last CommitMar 17, 2026

Actions

View Source View Plugin View on GitHub View README

Help us improve

Share bugs, ideas, or general feedback.

Stats

Actions

Help us improve

Share bugs, ideas, or general feedback.

Omni Routes

List routes

omni routes list --instance <id> --json
omni routes list --instance <id> --active --json
omni routes list --instance <id> --scope chat --json
omni routes list --instance <id> --scope user --json

Get route

omni routes get <routeId> --instance <id> --json

Create route

# Route a specific chat to an agent
omni routes create --instance <id> --scope chat --chat <chatId> --provider <providerId> --agent <agentId> --label "Support bot" --priority 10 --json

# Route all conversations from a person to an agent
omni routes create --instance <id> --scope user --person <personId> --provider <providerId> --agent <agentId> --json

# Create with gate and streaming options
omni routes create --instance <id> --scope chat --chat <chatId> --provider <providerId> --agent <agentId> --gate --gate-model <model> --gate-prompt "Only reply if relevant" --stream --timeout 60 --json

# Create as inactive (not yet active)
omni routes create --instance <id> --scope chat --chat <chatId> --provider <providerId> --agent <agentId> --inactive --json

Update route

omni routes update <routeId> --instance <id> --label "New label" --priority 20 --json
omni routes update <routeId> --instance <id> --active --json
omni routes update <routeId> --instance <id> --inactive --json
omni routes update <routeId> --instance <id> --agent <newAgentId> --timeout 120 --json
omni routes update <routeId> --instance <id> --no-gate --no-stream --json

Delete route

omni routes delete <routeId> --instance <id> --json

Test resolution

# Test which route would be applied for a given chat
omni routes test --instance <id> --chat <chatId> --json

# Test for a specific person
omni routes test --instance <id> --person <personId> --json

# Test for both chat and person
omni routes test --instance <id> --chat <chatId> --person <personId> --json

Metrics

omni routes metrics --json

Notes

Routes override instance-level agent config for matched chats or persons.
Scope "chat" targets a specific conversation; "user" targets all conversations from a person.
Higher --priority value wins when multiple routes could match.
--gate enables an LLM-based filter that decides whether the agent should reply.
--reply-filter-mode accepts all or filtered.
--agent-type accepts agent, team, or workflow (default: agent).