skills.rest
SeaBrad72
SeaBrad72COMMUNITY
Explore all skills by this author →
GitHub

verification

Run fresh test suites and conformance checks before commits.

1|Updated Jul 12, 2026
Search Tags:#testing#quality assurance#verification#sdlc#guardrails#confabulation-proofing
One-click install
npx skills add https://github.com/SeaBrad72/sparkwright --skill verification-seabrad72
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: verification
Source: https://github.com/SeaBrad72/sparkwright/tree/main/skills/verification
Command: npx skills add https://github.com/SeaBrad72/sparkwright --skill verification-seabrad72

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill prevents the common AI failure of claiming work is complete without actual proof, ensuring that every "done" status is backed by verifiable, fresh execution results.

Core Features & Use Cases

  • Evidence-Based Completion: Enforces the Iron Law that no completion claim can be made without running a fresh verification command and reading its output.
  • Confabulation-Proofing: Protects against subagent hallucinations by requiring verification on disk or via a clone dry-run rather than trusting internal reports.
  • Use Case: Before opening a pull request, use this skill to run your test suite and conformance checks, ensuring the current tree is truly green and not just relying on outdated memory or assumptions.

Quick Start

Run the verification skill to confirm the current project state by executing the primary test suite and validating the exit code.

Frequently Asked Questions about verification

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

FAQPage Schema
How do I prevent AI from claiming code is done without running tests?▼

Preventing AI confabulation requires enforcing an evidence-before-claims protocol that mandates fresh execution of test suites and conformance checks on disk before accepting any completion status.

What is evidence-based verification in the software development lifecycle?▼

Evidence-based verification is the practice of executing local commands and dry-run validations to ensure all project artifacts are validated on disk before opening a pull request or commit.

How do I validate my project tree is truly green before a pull request?▼

You validate the project tree by running the primary test suite, executing conformance checks, and reading the fresh command output to confirm a green state rather than relying on outdated memory.

Can I use dry-run validation to proof my work against subagent hallucinations?▼

Yes, you can use clone dry-run validation to protect against subagent hallucinations by requiring verification of artifacts on disk rather than trusting internal reports.

Do I need any dependencies to run conformance checks for confabulation-proofing?▼

No external dependencies are required; the skill operates independently across the development lifecycle using local command execution to enforce rigorous verification of software completion claims.

Related Skills

Explore More Software Engineering Skills→
spiralhousespiralhouse
5

baseline-check

Run RuboCop and RSpec checks against previous baselines to detect regressions.

Official
Intermediate
DYAI2025DYAI2025
3

verification-before-completion

Verify task status with fresh test, build, lint, and regression evidence.

Community
Basic
cwijayasundaracwijayasundara
3

harness-core

Automate build, test, deployment, and code review in CI/CD pipelines.

Community
Advanced
shekohexshekohex
9

verification-before-completion

Run verification commands and confirm output before claiming task completion.

Community
Basic