gaia-review

Assign probabilistic priors to independent Gaia claims in priors.py.

32|9|Updated Mar 2, 2026
One-click install
npx skills add https://github.com/SiliconEinstein/Gaia --skill gaia-review
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gaia-review
Source: https://github.com/SiliconEinstein/Gaia/tree/main/gaia/_skills/gaia-review
Command: npx skills add https://github.com/SiliconEinstein/Gaia --skill gaia-review

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Gaia review helps reviewers correctly assign probabilistic priors to independent Gaia claims so belief propagation produces results that match the reviewer’s intent.

Core Features & Use Cases

  • Prior-assignment workflow: turn the package’s reviewer intent into a complete priors.py using register_prior with non-empty justifications.
  • Role-aware prior guidance: identifies which claims need priors (independent/background/orphaned) and which must not (derived conclusions).
  • Iteration and publish readiness: supports a structured loop (brief → hole → edit priors.py → infer → interpret → gate) to reach an internally consistent, reviewable package.

Quick Start

Run gaia build check for the target package with --hole, then use the resulting Holes list to author priors.py via gaia author register-prior until the next gaia run infer and gate checks pass.

Frequently Asked Questions about gaia-review

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

FAQPage Schema
How do I assign probabilistic priors to independent claims in a knowledge graph?

Belief propagation requires priors only on independent, background, or orphaned claims; you must avoid assigning priors to derived conclusions to maintain a valid probabilistic reasoning chain.

What's the best way to fix incorrect priors.py during Gaia package formalization?

Iterate using a structured workflow: run a build check to find holes, edit priors.py with justified probability ranges, re-run gaia run infer, and validate the results against the gate.

Why does belief propagation produce posterior beliefs that don't match my intent?

Belief propagation yields mismatched posterior beliefs when your priors.py file is missing, incorrect, or out of alignment with the evidence and discrimination story. Assign correct priors to independent claims to resolve this.

When do I need to avoid assigning priors in a probabilistic reasoning workflow?

Avoid assigning priors to derived conclusions in your probabilistic reasoning workflow. Priors belong strictly on independent, background, or orphaned claims to prevent skewing belief propagation results.