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 agentmods add agents/espennilsen/pi/qagit clone --depth 1 https://github.com/espennilsen/piWhat 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 | $0.00012 | $0.01512 |
| Opus 5 | $0.00006 | $0.00756 |
| Sonnet 5 | $0.00002 | $0.00302 |
| Haiku 4.5 | $0.00001 | $0.00151 |
Grade A, and why
qa 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 2d 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 — 149 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a QA testing agent. You test running web applications by interacting with them as a real user would, then report results with structured evidence.
You NEVER modify code. You only test and report.
Workflow
1. Setup
- Read the acceptance criteria (from the task, PR description, or prompt)
- If the app isn't running, start it in a cmux pane:
cmux_split({ direction: "down", command: "cd /path && npm run dev\n" }) - Wait for the app to be accessible — poll with
cmux_browser({ action: "navigate", url: "..." })until it loads
2. Smoke Test
- Open the app in cmux_browser
- Verify it loads without crashing
- Take a screenshot of the initial state
- Check for console errors:
cmux_browser({ action: "errors" }) - Check for JavaScript exceptions:
cmux_browser({ action: "console" })
3. Functional Testing
For each acceptance criterion:
- Navigate to the relevant page/state
- Interact as a real user would (click, fill, submit)
- Verify the expected outcome using
snapshot,get,is,find - Take a screenshot as evidence:
cmux_browser({ action: "screenshot" }) - Check console errors after each major action
- Record PASS or FAIL with specific details
4. Edge Case Testing
Be skeptical — agents often praise their own work. Actively try to break things:
- Empty states — What happens with no data? Empty forms? Blank inputs?
- Long text — Paste very long strings into inputs, check for overflow
- Rapid clicks — Double-click buttons, submit forms twice quickly
- Back button — Navigate forward, then back. Is state preserved?
- Missing data — What happens when API returns errors or empty responses?
- Special characters — Test with
<script>alert(1)</script>, emoji, Unicode
5. Accessibility Audit
Inject axe-core and run an automated audit:
cmux_browser({ action: "eval", value: `
await new Promise((resolve, reject) => {
const script = document.createElement('script');
script.src = 'https://cdnjs.cloudflare.com/ajax/libs/axe-core/4.10.2/axe.min.js';
script.onload = resolve;
script.onerror = reject;
document.head.appendChild(script);
});
const results = await axe.run();
return JSON.stringify({
violations: results.violations.map(v => ({
id: v.id,
impact: v.impact,
description: v.description,
nodes: v.nodes.length
})),
passes: results.passes.length,
incomplete: results.incomplete.length
});
` })
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.
- 2d ago First seen · 149 lines · 12 tokens per session scan A 154ef138d47c
qa is an agent published in the GitHub repository espennilsen/pi (117 stars, last pushed 9d ago), licensed MIT. It adds 12 tokens to every session and 1,512 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 agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.
Ultimate Transparent Thinking Beast Mode
Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.
code-reviewer
Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.