How does the confidence scoring algorithm work?

The skill uses a weighted point system that awards points for passing tests (40), successful builds (20), high code coverage (20), UI screenshots (10), and zero linting errors (10).

Can this skill automatically approve my code changes?

Yes, if the generated confidence score meets the threshold (typically >= 95%), the skill can trigger an auto-approval to move a task from 'In Review' to 'Done'.

What is the primary purpose of the Proof-of-Work skill?

It generates verifiable evidence, such as screenshots and test reports, to provide concrete proof that an AI-driven task was completed correctly and meets quality standards.

Which tools are used for UI screenshot capture?

The skill provides patterns specifically designed for Playwright to capture full-page screenshots at desktop (1920x1080), mobile (375x667), and tablet (768x1024) resolutions.

Proof of Work Validation

Name: Proof of Work Validation
Author: MadAppGang

byMadAppGang

•

214

安全与测试

Generates verifiable proof artifacts and confidence scores to validate autonomous task completion and enable auto-approvals.

关于

The Proof-of-Work skill provides a structured framework for Claude to verify its own output through automated artifact generation and validation logic. By capturing multi-device screenshots using Playwright, parsing test results, and calculating weighted confidence scores, it ensures that every bug fix, feature, or UI change is backed by concrete evidence. This skill is essential for developers using autonomous agents who need to maintain high quality standards while streamlining the code review and deployment process.

主要功能

Multi-resolution screenshot capture for UI verification
Standardized markdown proof summary generation
Automated test result parsing and coverage reporting
214 GitHub stars
Weighted confidence scoring algorithm (0-100%)
Threshold-based task auto-approval logic

使用场景

Calculating deployment readiness based on test coverage and build status
Verifying UI consistency across desktop, mobile, and tablet viewports
Automating the validation of bug fixes before pull request submission

关于

主要功能

Multi-resolution screenshot capture for UI verification
Standardized markdown proof summary generation
Automated test result parsing and coverage reporting
214 GitHub stars
Weighted confidence scoring algorithm (0-100%)
Threshold-based task auto-approval logic

使用场景

Calculating deployment readiness based on test coverage and build status
Verifying UI consistency across desktop, mobile, and tablet viewports
Automating the validation of bug fixes before pull request submission