Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add MadAppGang/claude-code --skill proof-of-workgit clone --depth 1 https://github.com/MadAppGang/claude-codeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/madappgang/claude-code/proof-of-work)<a href="https://agentmods.dev/skills/madappgang/claude-code/proof-of-work"><img src="https://agentmods.dev/badge/skills/madappgang/claude-code/proof-of-work/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/madappgang/claude-code/proof-of-work"><img src="https://agentmods.dev/badge/skills/madappgang/claude-code/proof-of-work.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00024 | $0.01470 |
| Opus 5 | $0.00012 | $0.00735 |
| Sonnet 5 | $0.00005 | $0.00294 |
| Haiku 4.5 | $0.00002 | $0.00147 |
Grade A, and why
proof-of-work scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 235 lines — stays where its author put it; the contents beside it link to each section on GitHub.
plugin: autopilot updated: 2026-01-20
Proof-of-Work
Version: 0.1.0 Purpose: Generate validation artifacts for autonomous task completion Status: Phase 1
When to Use
Use this skill when you need to:
- Generate proof artifacts after task completion
- Capture screenshots for UI verification
- Parse and report test results
- Calculate confidence scores for task validation
- Determine if a task can be auto-approved
Overview
Proof-of-work is the mechanism that validates task completion. Every finished task must include verifiable artifacts that demonstrate the work was done correctly.
Proof Types by Task
Bug Fix Proof
| Artifact | Required | Purpose |
|---|---|---|
| Git diff | Yes | Show minimal, focused changes |
| Test results | Yes | All tests passing |
| Regression test | Yes | Specific test for the bug |
| Error log (before/after) | Optional | Visual evidence |
Feature Proof
| Artifact | Required | Purpose |
|---|---|---|
| Screenshots | Yes | Visual verification |
| Test results | Yes | Functionality works |
| Coverage report | Yes | >= 80% coverage |
| Build output | Yes | Builds successfully |
| Deployment URL | Optional | Live demo |
UI Change Proof
| Artifact | Required | Purpose |
|---|---|---|
| Desktop screenshot | Yes | 1920x1080 view |
| Mobile screenshot | Yes | 375x667 view |
| Tablet screenshot | Yes | 768x1024 view |
| Accessibility score | Yes | >= 80 Lighthouse |
| Visual regression | Optional | BackstopJS diff |
Screenshot Capture
Playwright Pattern:
import { chromium } from 'playwright';
async function captureScreenshots(url: string, outputDir: string) {
const browser = await chromium.launch({ headless: true });
const context = await browser.newContext();
const page = await context.newPage();
// Desktop
await page.setViewportSize({ width: 1920, height: 1080 });
await page.goto(url);
await page.waitForLoadState('networkidle');
await page.screenshot({
path: `${outputDir}/desktop.png`,
fullPage: true,
});
// Mobile
await page.setViewportSize({ width: 375, height: 667 });
await page.goto(url);
await page.waitForLoadState('networkidle');
await page.screenshot({
path: `${outputDir}/mobile.png`,
fullPage: true,
});
// Tablet
await page.setViewportSize({ width: 768, height: 1024 });
await page.goto(url);
await page.waitForLoadState('networkidle');
await page.screenshot({
path: `${outputDir}/tablet.png`,
fullPage: true,
});
await browser.close();
}
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 235 lines · 24 tokens per session scan A eab6e200ef95
proof-of-work is a skill published in the GitHub repository MadAppGang/claude-code (279 stars, last pushed 5mo ago), licensed MIT. It adds 24 tokens to every session and 1,470 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
debugging-failures
Diagnose a failing Ginkgo suite as an agent — always run with --json-report into a predictable temp/gitignored location, read the terminal verdict line, then use jq to extract structured failure details (name, message, file:line, panic value, captured logs). Covers the panicked-vs-failed trap, panic locations pointing…
smiles-validation
Strict SMILES validation, structural comparison, and modification verification. Catches invalid LLM-generated molecules.
hypothesis-testing
Applies the scientific method to debugging by helping users form specific, testable hypotheses, design targeted experiments, and systematically confirm or reject theories to find root causes. Use when a user says their code isn't working, they're getting an error, something broke, they want to troubleshoot a bug, or…
doris-debug-deployment
Use for Doris FE/BE startup failures, port conflicts, prioritynetworks misrouting, metadir corruption, and ADD/DROP BACKEND issues.
app-platform-troubleshooting
Debug running App Platform applications by accessing containers, analyzing logs, running diagnostics, and applying fixes. Use when apps fail to deploy, crash at runtime, have connectivity issues, or need performance diagnosis.
k8s-troubleshoot
Debug Kubernetes pods, nodes, and workloads. Use when pods are failing, containers crash, nodes are unhealthy, or users mention debugging, troubleshooting, or diagnosing Kubernetes issues.