agent-dispatcher

# agent-dispatcher

[](https://github.com/nahid-sparktales/agent-dispatcher/actions/workflows/ci.yml)
[](https://github.com/nahid-sparktales/agent-dispatcher/actions/workflows/security.yml)

**Give each Claude Code or Codex task the specialist, context, and skills it needs.**

Agent Dispatcher routes your request to a focused role, loads relevant guidance, and defines
what evidence will count as done. Ask it to debug a failure, design an interface, review a
change, or research a decision. For substantial code tasks, it maintains local project facts
and code relationships, reuses valid parsing results, and selects a bounded context packet
for the current role. Small, known edits can take the direct path.

<!-- counts:start -->**27 roles · 79 local skills · 31 external skills · 8 recipes · 19 MCP servers · 50 detection signals**<!-- counts:end -->

[How it works](#how-it-works) · [Quick start](#quick-start) · [Usage](#usage) ·
[Project mapping](#keep-a-project-map-current) · [Catalog](#catalog) ·
[Documentation](#documentation) · [Contributing](#contributing)

## Why use it?

- **Match the method to the task.** A debugger reproduces the failure; a reviewer evaluates the
  change; a designer works through the interface and interaction.
- **Load context as needed.** Roles select a small set of skills, project files, and available
  tools. Full skill instructions are read only when selected.
- **Reuse project knowledge.** Source-linked facts and a structural graph help locate relevant
  code. A private incremental cache lets later preparation reuse unchanged sources and the resolved graph.
- **Make verification explicit.** The role defines the evidence needed and reports which checks
  actually ran, what passed, and what remains unverified.
- **Respect the task's edit scope.