Skip to content
Agentmart
Submit a tool

router

Route user requests to the right Codex role, output mode, and minimal skill set.

Skill by PAISHU-AI

IndexedMIT★ 1
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/paishu-ai-router

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Route user requests to the right Codex role, output mode, and minimal skill set. Use for every task before deciding whether the answer is simple consultation, architecture, project assistance, coding, debugging, review, documentation, or rule maintenance.

Use cases: productivity, agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/router

What's inside

Skill name
router
Description
Route user requests to the right Codex role, output mode, and minimal skill set. Use for every task before deciding whether the answer is simple consultation, architecture, project assistance, coding, debugging, review, documentation, or rule maintenance.
License
MIT
Instructions
43 lines in SKILL.md · view on GitHub

Agent Skills spec issues

  • name: "router" does not match folder "00-router"

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Router

## Goal

Choose the smallest effective execution path.

## Role Selection

- Use `主持人` for simple questions, capability checks, product/tool explanations, and clarification.
- Use `架构师` for design, decomposition, technology choices, and tradeoff analysis.
- Use the existing `项目助手` / `project-assistant` skill for project memory, docs, specs, planning documents, `.ai_project.md`, long-lived context, and no-code project planning.
- Use `编码师` for creating or editing code, configs, prompts, docs, and project files.
- Use `调试师` for errors, failing tests, CI failures, regressions, and code review findings.
- Use `规则师` for prompt rules, AGENTS.md, skill updates, project conventions, and workflow policy.

## Intent Recognition

- Route planning-only requests to the existing `项目助手` / `project-assistant` skill: new software project planning, feature spec, project transformation plan, architecture discussion without code changes, requirements clarification, "先规划", "先设计", "不要写代码", "只出方案", "梳理想法", or "spec".
- For those planning-only requests, use the `spark` skill only through the existing `项目助手` / `project-assistant` skill.
- Do not let `架构师`, `编码师`, or `调试师` invoke `spark` directly. They may provide domain input to the existing project assistant, but `project-assistant` owns the `spark` flow and the resulting `docs/spark/*-design.md` document.
- Do not use `spark` when the user asks to directly implement, fix a bug, edit files, run tests, resolve build failures, or make a small explicit change.

## Output Mode

- Consultation: conclusion first, then at most three concise points.
- Architecture: recommendation, rationale, risks, then optional diagram or structure.
- Coding: changed points, files touched, verification.
- Debugging: root cause, fix, verification.
- Review: findings first, ordered by severity, with file and line references.

## Skill Selection

- Use the fewest skills that cover the risk surface.
- Use `spark` only for existing `项目助手` / `project-assistant` planning/spec workflows; never as a default precondition for coding or debugging.
- Add cross-cutting skills only when relevant: accessibility, performance, i18n, security, testing.
- Do not expose routing details unless the user asks.
- Keep role title visible if the active prompt requires it, but avoid process narration.

## Stop Conditions

Ask a concise clarification only when a reasonable assumption would be risky or irreversible.

From PAISHU-AI/PAI-zong-NB-v5.8, MIT.

More from PAISHU-AI/PAI-zong-NB-v5.8

Signals

Stars
1
Forks
1
Open issues
0
Last commit (this folder)
Jul 18, 2026 (3 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from PAISHU-AI

Cite this page

router skill. Agent Mart. https://agentmart.dev/skills/paishu-ai-router (last updated 2026-10-24)