skills.rest
gaiastudio-ai
gaiastudio-aiOFFICIAL
Explore all skills by this author →
GitHub

gaia-ground-truth-management

Manage and reconcile per‑agent ground-truth sidecar files and dual-refresh lifecycle metadata.

2|Updated Apr 15, 2026
Search Tags:#archival#data-integrity#brownfield#validator#token-budget#ground-truth#dual-refresh
One-click install
npx skills add https://github.com/gaiastudio-ai/gaia-public --skill gaia-ground-truth-management
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: gaia-ground-truth-management
Source: https://github.com/gaiastudio-ai/gaia-public/tree/main/plugins/gaia/skills/gaia-ground-truth-management
Command: npx skills add https://github.com/gaiastudio-ai/gaia-public --skill gaia-ground-truth-management

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ground-truth management is essential for validator workflows to maintain an authoritative record of verified facts and ensure updates are reconciled with filesystem state. This skill provides a deterministic lifecycle for per-agent ground-truth data, supporting entry structuring, incremental and full refresh strategies, conflict resolution, archival, token-budget governance, and brownfield seeding.

Core Features & Use Cases

  • Ground-truth rewrite path with deterministic operations over the per-agent ground-truth.md sidecar file.
  • Dual-refresh lifecycle ensuring both runtime sidecar and committed seed stay in sync.
  • Conflict resolution rules for factual conflicts, missing sources, and new facts; includes changelog notes.
  • Token-budget monitoring and archival guidance to keep within limits.
  • Brownfield extraction seeds initial ground truth from existing artifacts.

Quick Start

Seed and initialize the ground-truth management workflow for the project using default settings.

Frequently Asked Questions about gaia-ground-truth-management

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

FAQPage Schema
How do I manage ground-truth data for validator workflows?▼

Ground-truth data for validator workflows is managed through deterministic operations over a per-agent sidecar file. This lifecycle reconciles verified facts with filesystem state to ensure artifact integrity.

What is the dual-refresh lifecycle for ground-truth files?▼

The dual-refresh lifecycle synchronizes the runtime sidecar file with the committed seed. This ensures both runtime state and committed truth remain consistent across project cycles.

How do I extract brownfield ground truth from existing project artifacts?▼

Brownfield extraction seeds initial ground truth by analyzing and pulling verified facts from existing project artifacts. This establishes a baseline sidecar file before applying the dual-refresh lifecycle.

How are factual conflicts and missing sources resolved in ground-truth files?▼

Factual conflicts, missing sources, and new facts are resolved using specific conflict resolution rules. This process enforces the empty-seed invariant and generates changelog notes for transparency.

How do I monitor token budgets for ground-truth sidecar files?▼

Token-budget monitoring tracks the size of ground-truth sidecar files to keep them within limits. Archival guidance is applied when files approach the token budget to maintain operational efficiency.

Does ground-truth management enforce deterministic section loading via ADR-041?▼

Ground-truth management enforces deterministic section loading via ADR-041 runtime rules. This ensures stable, predictable loading of ground-truth sections during validator execution.

Related Skills

Explore More Data & Analytics Skills→
gaiastudio-aigaiastudio-ai
2

gaia-refresh-ground-truth

Rescan the filesystem and regenerate a structured diff report against the last ground-truth snapshot.

Official
Advanced
flashinfer-aiflashinfer-ai
264

add-reference-tests

Generate and validate pytest tests for flashinfer_trace reference implementations against FlashInfer or SGLang ground truth.

Official
Advanced
gaiastudio-aigaiastudio-ai
2

gaia-memory-management

Read and write GAIA agent memory sidecars with ADR-016 formatting.

Official
Advanced
EurecaMomentEurecaMoment

benchclaw-stage4-gt-kinship-analysis

Analyze and normalize ground truth relationships for benchmark template generation.

Official
Advanced