task-taxonomy

Classify research tasks into seven primitive categories for model routing.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/Biajin-PKU/research-harness --skill task-taxonomy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: task-taxonomy
Source: https://github.com/Biajin-PKU/research-harness/tree/main/codex-skills/task-taxonomy
Command: npx skills add https://github.com/Biajin-PKU/research-harness --skill task-taxonomy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The Task Taxonomy skill helps researchers and AI agents decide which model path and primitives to apply for a given research task, enabling efficient routing and consistent results.

Core Features & Use Cases

  • Categorizes primitives into seven categories: Retrieval, Comprehension, Extraction, Analysis, Synthesis, Generation, and Verification, each with specific primitives and routing implications.
  • Provides concrete use cases across tasks like literature reviews, data extraction, hypothesis generation, and manuscript drafting to guide tool selection and sequencing.
  • Serves as a reference implementation for routing policies in MCP servers, enabling predictable behavior and auditability.

Quick Start

Provide a research task description and have the system classify it into a primitive category with routing guidance.

Frequently Asked Questions about task-taxonomy

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I classify research tasks for model routing?

The seven primitive categories used for task classification are Retrieval, Comprehension, Extraction, Analysis, Synthesis, Generation, and Verification, each mapping to specific research actions and routing paths.

What is the best way to route AI agents for literature reviews and experiment planning?

Routing AI agents for literature reviews and experiment planning uses a task taxonomy to map these activities to defined primitives, ensuring predictable model selection and consistent tool sequencing.

Does task taxonomy support MCP-based routing rules?

Yes, task taxonomy implements MCP-based routing rules to serve as a reference for routing policies in MCP servers, enabling predictable behavior and auditability for model selection.

Can I use this classification system for proposal drafting and evidence synthesis?

Yes, you can use this classification system for proposal drafting and evidence synthesis, as it applies these research tasks to decide appropriate primitives and routing paths for consistent results.

What are the limitations of using a seven-category taxonomy for research task routing?

A limitation of using a seven-category taxonomy for research task routing is that complex, multi-disciplinary queries may overlap categories, requiring manual disambiguation to ensure accurate model and primitive selection.