task-router

task-router is a skill for Claude Code, Codex from Ozzeron/prompt-pack. It costs 24 tokens per session (2,099 once invoked), scanned A, original, MIT.

A task-routing and coordination skill for choosing the right specialist instructions for a request. It can also decide whether several tasks should be handled by separate subagents, meaning additional agents working on parts of the job.

In plain words
What is it for?
Use it for non-trivial tasks, work spanning several disciplines, and deciding which specialist should handle backend, frontend, review, or other parts of a request.
Why use it?
It prevents complex requests from being handled by the wrong specialist or from being split up without a clear reason. It provides a consistent way to coordinate multi-step work.

Skill for Claude CodeCodex

Which agent this was written for is unclear — built for openclaw. Also seen: mentions CLAUDE.md; mentions subagents; mentions AGENTS.md.

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/ozzeron/prompt-pack/task-router
Any agent
npx skills add Ozzeron/prompt-pack --skill task-router
Clone the repo
git clone --depth 1 https://github.com/Ozzeron/prompt-pack

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for task-router

README.md
[![agentmods](https://agentmods.dev/badge/skills/ozzeron/prompt-pack/task-router.svg)](https://agentmods.dev/skills/ozzeron/prompt-pack/task-router)
Your own site
<a href="https://agentmods.dev/skills/ozzeron/prompt-pack/task-router"><img src="https://agentmods.dev/badge/skills/ozzeron/prompt-pack/task-router.svg" alt="Measured on agentmods" height="20"></a>
Per session 24 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,099 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00024 $0.02099
Opus 5 $0.00012 $0.01050
Sonnet 5 $0.00005 $0.00420
Haiku 4.5 $0.00002 $0.00210

Measured 6d ago against content hash c6fd36ca73af, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

task-router scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

prompts/meta/task-router/SKILL.md · 156 lines

How it starts

The opening of the file, as written. The whole thing — 156 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Task Router

You are the orchestrator. The user gives you work; your job is to decide which specialist role(s) handle it, whether to do it inline or delegate to a subagent, and how to combine the results.

You do not write the deliverable yourself when a specialist exists for it. You coordinate.

When to use

  • Any non-trivial multi-step request
  • Requests that span multiple disciplines (e.g. backend + frontend + review)
  • Requests where the right role is not obvious from the user's wording
  • Before spawning a subagent — check the routing table first

For one-off trivial requests (a single edit, a quick question), skip routing and answer directly.

Routing table

Only skills that actually exist in this pack are routed here. As new skills land, this table grows. Don't route to a skill that isn't in prompts/.

User intent Primary skill Subagent? Notes
Build a backend endpoint / API route architecture/backend-api Optional Spawn if work spans multiple files
Build a frontend feature / page architecture/frontend-feature Optional Inherits ui-designer for the UI portion
Design a UI / new screen interface/ui-designer Optional Ask creativity level first
Design new tables / data model architecture/database-schema Optional Add postgres-supabase if Supabase/Postgres
Write a DB migration architecture/database-migrations Optional Pair with database-schema for design changes
Supabase RLS / auth / migration workflow architecture/postgres-supabase Optional Inherits database-schema and database-migrations
Audit existing frontend codebase review/frontend-audit Yes (fork) Audits are read-heavy; isolate context
Audit / review whole project (no diff) review/repo-audit Yes (fork) Sample across areas; honest coverage section in output. Use when the user asks for a project-wide review without pointing at a diff.
Code review on a diff or PR review/code-review Optional Subagent if diff is large
Security review on a diff or module review/security-review Yes Always isolate; complements code-review
Review schema, query, or migration review/database-review Optional Subagent if multiple tables involved
Find code duplication / DRY audit review/duplication-audit Optional Grep-first; pairs with architecture/refactor-planner for execution
Debug a failing test or bug review/debugger Yes (fork) Hypothesis-first; bring context, isolate steps
Plan a refactor / migration architecture/refactor-planner Yes Plans need uninterrupted thinking; outputs plan, not diff
Write tests for existing code delivery/test-writer Optional Behaviour over implementation; AAA structure
Write or update docs (README, ADR, doc comments, API docs) delivery/doc-writer Optional Grounded in actual code, never auto-publishes
Write AGENTS.md / CLAUDE.md / .cursorrules / agent instructions delivery/ai-agent-docs Optional Specialised version of doc-writer for AI-readable docs
Write or modify Dockerfile / docker-compose / containerize / .dockerignore infra/docker Optional Subagent if multi-service compose or multi-stage refactor
Wrap up / hand off completed work delivery/handoff No Inline at end of any coding task

Read the full file on GitHub · 156 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 6d ago First seen · 156 lines · 24 tokens per session scan A c6fd36ca73af

Subscribe to this mod's changes

task-router is a skill published in the GitHub repository Ozzeron/prompt-pack (8 stars, last pushed 1mo ago), licensed MIT. It adds 24 tokens to every session and 2,099 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.