skills.rest
EquiStamp
EquiStampOFFICIAL
Explore all skills by this author →
WebsiteGitHub

evaluation-building

Translate a validated Phase 3 evaluation design into a Phase 4 Build Plan.

Updated Aug 27, 2026
Search Tags:#documentation#evaluation#handoff#inspect#phase-4#instrument#build-plan
One-click install
npx skills add https://github.com/EquiStamp/evaluating-evaluations --skill evaluation-building
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: evaluation-building
Source: https://github.com/EquiStamp/evaluating-evaluations/tree/main/.claude/skills/evaluation-building
Command: npx skills add https://github.com/EquiStamp/evaluating-evaluations --skill evaluation-building

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Transforms a validated evaluation design into a concrete, reproducible Phase 4 build plan and implementation guide, bridging design to deployment.

Core Features & Use Cases

  • Create a comprehensive Build Plan with architecture, item-set design, scoring pipeline, and validation steps for an Inspect-based evaluation.
  • Provide phase-4 implementation guidance including docker configuration, code structure, and deployment considerations.
  • Support end-to-end handoff with documentation and testable artifacts to enable a non-builder to reproduce results.

Quick Start

Provide a Phase 3 design and I will generate a complete Phase 4 build plan and guidance.

Frequently Asked Questions about evaluation-building

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

FAQPage Schema
How do I generate an Inspect-based build plan from a validated evaluation design?▼

An AI safety evaluation build plan details architecture, item-generation strategy, scoring rules, logging, and Inspect task guidance. It bridges validated designs to reproducible deployment.

How do I create a scoring pipeline and item-set design for an AI safety evaluation?▼

Use the translated Phase 4 build plan to configure your scoring pipeline and item-set design. It provides scoring rules, item-generation strategy, and validation steps for Inspect-based pipelines.

Do I need a Phase 3 evaluation design to create a Phase 4 build plan?▼

Yes, a validated Phase 3 evaluation design is required as input. The system uses it to produce the Phase 4 build plan, implementation guide, and testable artifacts for handoff.

Can I use this approach to generate docker configuration and code structure for evaluation handoff?▼

Yes, the approach provides phase-4 implementation guidance including docker configuration and code structure. This enables end-to-end handoff with documentation and testable artifacts for non-builders.

What's the best way to document infohazard controls and validation steps for an Inspect evaluation?▼

Document infohazard controls and validation steps within a generated build plan. It applies these controls to Inspect-based pipelines alongside architecture, repository structure, and scoring rules.

Related Skills

Explore More Software Engineering Skills→
nuggetswisenuggetswise

writing-product-strategy

Generate a 3-5 year product strategy document with vision, pillars, roadmap, and moats.

Community
Advanced
MohammedHTahirMohammedHTahir

deployments-cicd

Guide Vercel deployments and CI/CD across GitHub Actions, GitLab CI, and Bitbucket Pipelines.

Community
Advanced
SylvainChabaudSylvainChabaud
10

gate-check

Inspect repository structure and report missing artifacts for CI/CD gates 0 through 5.

Community
Intermediate
GTC6244GTC6244

plan-eng-review

Plan and review engineering architecture and execution plans.

Community
Advanced