proof-verify

Enforces a plan-based verification workflow using frozen specifications and independent agents.

140|23|Updated Mar 28, 2026
One-click install
npx skills add https://github.com/AnastasiyaW/codex-claude-code-config --skill proof-verify-anastasiyaw
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: proof-verify
Source: https://github.com/AnastasiyaW/codex-claude-code-config/tree/main/skills/development/proof-verify
Command: npx skills add https://github.com/AnastasiyaW/codex-claude-code-config --skill proof-verify-anastasiyaw

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill solves the issue of confirmation bias and incomplete testing by enforcing a strict separation between the builder of a feature and the verifier, ensuring that code is validated against frozen acceptance criteria rather than the builder's own assumptions.

Core Features & Use Cases

  • Plan-Based Verification: Requires a frozen plan with testable acceptance criteria before any code is written.
  • Independent Review: Mandates that a fresh, isolated agent verifies the implementation without access to the builder's progress notes.
  • Use Case: Use this when implementing complex features or security-sensitive logic where you need to guarantee that the final output strictly adheres to a pre-defined specification.

Quick Start

Initiate the proof-verify workflow by creating a .proof/PLAN.md file in your project root and spawning an independent agent to validate your implementation against the defined criteria.

Frequently Asked Questions about proof-verify

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

FAQPage Schema
How do I enforce spec-driven verification for code against predefined acceptance criteria?

Spec-driven verification is enforced by isolating the builder from the review process through a frozen plan file and a fresh-context agent that independently evaluates the implementation against testable acceptance criteria.

How does independent agent code review prevent confirmation bias?

Independent agent code review prevents confirmation bias by verifying implementations in an isolated context without access to the builder's progress notes, ensuring validation strictly against frozen specifications rather than builder assumptions.

What's the best way to start a plan-based verification workflow for security-sensitive logic?

Start plan-based verification by creating a .proof/PLAN.md file containing testable acceptance criteria in your project root, then spawn an independent agent to validate the implementation against that frozen specification.

Do I need structured plan files before writing code to use spec-driven quality assurance?

Yes, structured plan files with testable acceptance criteria are required before code implementation to maintain high-integrity development cycles and enable independent quality assurance verification.

Why does independent verification require isolating the builder from the verification process?

Independent verification requires builder isolation to prevent the builder's own assumptions from influencing the validation, ensuring code is strictly validated against frozen acceptance criteria for high-integrity development.

When should I use plan-based verification instead of standard testing workflows?

Use plan-based verification when implementing complex features or security-sensitive logic where you need to guarantee the final output strictly adheres to a pre-defined specification without confirmation bias.