verify-agent-work

Reproduce autonomous coding agent output and inspect diffs in isolated Git checkouts.

7|23|Updated Aug 9, 2015
One-click install
npx skills add https://github.com/gamedevpl/www.gamedev.pl --skill verify-agent-work
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: verify-agent-work
Source: https://github.com/gamedevpl/www.gamedev.pl/tree/main/.claude/skills/verify-agent-work
Command: npx skills add https://github.com/gamedevpl/www.gamedev.pl --skill verify-agent-work

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents unverified or attacker-influenced autonomous coding work from being trusted or merged into a production codebase.

Core Features & Use Cases

  • Isolated Verification: Verify agent changes in a throwaway clone or worktree without risking active development work.
  • Comprehensive Validation: Run the project's real dependency, type-checking, linting, test, build, CI, deployment, and runtime gates.
  • Security and Scope Review: Inspect diffs for prompt injection, secrets, supply-chain changes, weakened assertions, configuration errors, and invariant violations.
  • Runtime Game Testing: Exercise agent-authored browser games with deterministic clock and canvas-pixel checks when static validation is insufficient.
  • Use Case: When an autonomous agent submits a pull request, independently reproduce its claims, verify the complete assembled application, inspect persistence and deployed configuration, and reject failures or scope creep before merging.

Quick Start

Use the verify-agent-work skill to independently validate the autonomous agent's changes in an isolated checkout and report only the failures and evidence you observe.

Frequently Asked Questions about verify-agent-work

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

FAQPage Schema
How do I verify autonomous agent code changes before merging?

To verify autonomous agent code changes before merging, reproduce the agent's behavior in an isolated Git checkout and inspect its diff for scope creep, security issues, and invariant violations before running your project's validation gates.

What security validation should I run on agent-generated pull requests?

Security validation for agent-generated pull requests should inspect diffs for prompt injection, leaked secrets, supply-chain changes, weakened assertions, and configuration errors to prevent attacker-influenced code from entering production.

Can I test agent-authored browser games with deterministic canvas checks?

Yes, you can test agent-authored browser games with deterministic clock and canvas-pixel checks when static validation is insufficient, exercising runtime behavior to verify the complete assembled application.

How do I isolate code verification for parallel-agent workflows?

You isolate code verification for parallel-agent workflows by using a throwaway clone or Git worktree, ensuring dependency consistency checks and real CI gates run without risking your active development work.

What's the best way to run CI gates on autonomous coding agent output?

The best way to run CI gates on autonomous coding agent output is to independently reproduce its claims in an isolated checkout, executing the project's actual type-checking, linting, tests, builds, and deployment checks to reject failures.