research-harness

Route general prompts to the most specific Research Harness workflow.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Research Harness provides a unified entry point that routes general requests to the appropriate research workflow, coordinating topic initialization, literature search, citation tracing, claim extraction, gap analysis, evidence gating, section drafting, paper verification, and provenance review.

Core Features & Use Cases

  • Routes requests to the most specific workflow based on topic and task.
  • Maintains persistent state, typed primitives, and stage-gated progression to ensure auditability and continuity.
  • Supports dependency ordering and safe fallbacks when tools are unavailable.

Quick Start

Provide a general prompt and let Research Harness route it to the appropriate workflow.

Frequently Asked Questions about research-harness

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

FAQPage Schema
How do I orchestrate end-to-end research workflows from a general prompt?

Research workflow orchestration routes general prompts to the most specific workflow, coordinating topic bootstrap, literature search, citation tracing, gap analysis, evidence gating, drafting, verification, and provenance review. It uses frontmatter-based discovery and explicit routing rules to manage the entire process.

What is research workflow routing and how does it handle task selection?

Research workflow routing directs user requests to the appropriate research workflow based on the topic and task. It applies safe defaults and fallbacks when specific tools are unavailable, ensuring continuous progression across stage-gated research tasks like claim extraction and mapping.

How do I manage persistent state and provenance across multiple research stages?

Persistent state and provenance are maintained through typed primitives and stage-gated progression. This ensures auditability and continuity across dependency-ordered research stages, from initial topic bootstrap to final paper verification and provenance review.

Can I use research orchestration when specific literature search tools are unavailable?

Research orchestration supports dependency ordering and safe fallbacks when specific tools are unavailable. It applies safe defaults during workflow routing to ensure that literature search, mapping, and citation tracing tasks can proceed without interruption.

What's the best way to extract claims and analyze gaps in academic literature?

Claim extraction and gap analysis are executed by routing general prompts to specialized research workflows. The orchestration harness coordinates these tasks alongside literature search and evidence gating to identify research gaps and extract verifiable claims systematically.