verify

Delegate build, test, lint, diff review, and readability checks to an implementation-verifier agent.

Updated Nov 4, 2022
One-click install
npx skills add https://github.com/CityBear3/dotfiles --skill verify-citybear3
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: verify
Source: https://github.com/CityBear3/dotfiles/tree/main/claude/skills/verify
Command: npx skills add https://github.com/CityBear3/dotfiles --skill verify-citybear3

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prevents premature success or completion claims by requiring fresh, command-backed verification after an implementation plan finishes.

Core Features & Use Cases

  • Formal post-implementation verification: Delegates to the implementation-verifier agent to run build, test, lint, diff review, and readability checks in a single gated phase.
  • Iron Law enforcement: Requires running verification commands in the current session and validating exit codes and outputs before expressing satisfaction.
  • Autonomous workflow transition: Automatically moves from verification to review when checks pass, and returns to execute-plan when they fail.

Quick Start

Use the verify skill after your /execute-plan run by instructing the AI to run /verify to produce fresh evidence-backed verification results.

Frequently Asked Questions about verify

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

FAQPage Schema
How do I verify code changes before claiming an implementation task is complete?

To verify code changes before claiming completion, run fresh build, test, and lint commands in the current session to inspect outputs and validate exit codes. This enforces evidence-backed confirmation of software delivery outcomes before transitioning to review.

What is post-implementation verification in an autonomous software delivery workflow?

Post-implementation verification is a gated phase that delegates diff review and readability checks to an implementation-verifier agent. It requires fresh command execution and output inspection to prevent premature success claims before transitioning to review or returning to execute-plan.

How do I prevent false completion claims when using autonomous agents for software testing?

To prevent false completion claims during software testing, enforce an Iron Law requiring fresh command execution and explicit exit-code failure handling within the same session. This ensures evidence-grounded status reporting before any done status is accepted.

Does the verify skill require dependencies or external components to run diff reviews?

No, the verify skill operates with no dependencies and no components. It delegates build, test, lint, diff review, and readability checks directly to an implementation-verifier agent within your existing software delivery workflow.

What happens when linting or build checks fail during the verification process?

When linting or build checks fail during verification, the workflow automatically returns to execute-plan to address the errors. It only transitions to review when all executed commands pass their exit-code validations and produce fresh evidence.

When do I need to run formal verification checks after executing my implementation plan?

You need to run formal verification checks immediately after your execute-plan run finishes. This produces fresh, command-backed evidence from build, test, and lint outputs required to confirm outcomes within the same session before moving to review.