ia-verification-before-completion

Require fresh verification evidence before reporting task completion.

30|3|Updated Feb 22, 2026

Search Tags:#code review #verification #scope control #agent workflows #release readiness #test discipline #CI evidence

Ready to run

npx skills add https://github.com/iliaal/whetstone --skill ia-verification-before-completion

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: ia-verification-before-completion
Source: https://github.com/iliaal/whetstone/tree/main/plugins/whetstone/skills/ia-verification-before-completion
Command: npx skills add https://github.com/iliaal/whetstone --skill ia-verification-before-completion

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill prevents AI coding agents from declaring work complete without fresh evidence, reducing false confidence and missed failures.

Core Features & Use Cases

Verification gate before claims: Requires running verification immediately before any completion statement such as “tests pass,” “bug fixed,” or “ready to merge.”
Pre-edit scope confirmation: Adds a spatial scope gate for ambiguous requests like “migrate my project” to avoid accidental cross-repo blast radius.
Evidence-based completion reporting: Forces a structured completion report including what was verified and the exact verification evidence.

Quick Start

Ask the AI to verify the change with the appropriate build/test/security commands and only then report completion with a structured verification evidence section.

Frequently Asked Questions about ia-verification-before-completion

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

FAQPage Schema

How do I stop AI coding agents from claiming work is complete without running tests?▼

To stop false completion claims, enforce a verification gate requiring fresh test evidence before any status signaling. This prevents AI agents from declaring a bug fixed or ready to merge without current proof of passing tests or successful builds.

Why does my AI agent claim a migration is finished without verifying the build?▼

AI agents claim migrations are finished without verification due to missing pre-verification state checks. Enforcing a stepwise verification pipeline that stops on first failure ensures completion reports include exact evidence and risk notes before signaling success.

What's the best way to control scope when asking an AI agent to migrate a project?▼

The best way to control scope during ambiguous edits like project migrations is applying a spatial scope-confirmation gate. This prevents accidental cross-repo blast radius by requiring explicit scope confirmation before the agent executes broad modifications.

How do I enforce CI evidence collection before declaring release readiness?▼

Enforce CI evidence collection for release readiness by mandating an explicit pre-verification state check immediately before completion statements. This structured pipeline requires running appropriate build and test commands, stopping on first failure, and reporting verification evidence.

Does this verification workflow apply to delegated subagent coding and refactoring tasks?▼

Yes, this verification workflow applies directly to delegated subagent coding, infrastructure refactoring, and migrations. It enforces evidence-based completion reporting with risk notes across all agent workflows where status-signaling would otherwise happen without current proof.

What is a pre-verification state check for AI agent workflows?▼

A pre-verification state check for AI agent workflows is an explicit gate requiring fresh verification evidence before any completion claim. It ensures agents run appropriate build, test, or security commands and generate a structured completion report before signaling success.