gaia-playbook

Diagnose GAIA trajectory gaps and select intervention patterns for score improvements.

415|49|Updated Apr 30, 2026
One-click install
npx skills add https://github.com/Darwin-Agent/HarnessX --skill gaia-playbook
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gaia-playbook
Source: https://github.com/Darwin-Agent/HarnessX/tree/main/recipe/gaia_evolver/skills/gaia-playbook
Command: npx skills add https://github.com/Darwin-Agent/HarnessX --skill gaia-playbook

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

GAIA benchmark guidance to diagnose trajectory gaps, map failure modes to actionable interventions, and streamline tool-spec authoring for performance improvement.

Core Features & Use Cases

  • Maps GAIA failure modes (A-H) to concrete tooling patterns and harness configurations.
  • Provides templates and best practices for drafting TOOL_SPEC.md and coordinating with reference materials.
  • Supports teams in planning, experimentation, and score improvement across multi-hop GAIA tasks.

Quick Start

Identify your GAIA trajectory gaps and draft a TOOL_SPEC.md to address the most critical capability missing.

Frequently Asked Questions about gaia-playbook

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

FAQPage Schema
How do I diagnose GAIA benchmark trajectory gaps to improve my agent's score?

To diagnose GAIA trajectory gaps, map your agent's failure modes to actionable interventions using a structured playbook. This identifies missing capabilities in multi-hop tasks and provides specific tooling patterns to drive score improvement.

What is the best way to structure a TOOL_SPEC.md for GAIA multi-hop tasks?

The best way to structure a TOOL_SPEC.md for GAIA tasks is using frontmatter-driven templates. These templates coordinate with reference materials to standardize tool-shape design and ensure consistent harness configurations.

How do GAIA failure modes map to specific tooling patterns and harness configurations?

GAIA failure modes map to specific tooling patterns through categorized trajectory analysis. By identifying distinct failure clusters, you apply targeted HarnessX levers like config, control, action, and instruction adjustments to correct agent behavior.

Can I use HarnessX levers to improve agent performance across different GAIA clusters?

Yes, you can use HarnessX levers to improve agent performance across GAIA clusters. Adjusting config, control, action, and instruction parameters provides targeted interventions that address specific trajectory gaps and drive score improvements.

Why does my GAIA benchmark agent fail on multi-hop tasks despite having the right tools?

Your GAIA agent may fail multi-hop tasks due to unmapped failure modes in trajectory analysis. Even with correct tools, misaligned instruction patterns or improper harness configurations prevent successful multi-step reasoning and task completion.