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 instructions/microsoft/playwright/copilot-instructionsgit clone --depth 1 https://github.com/microsoft/playwrightWhat 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 | $0.00121 | $0.00121 |
| Opus 5 | $0.00060 | $0.00060 |
| Sonnet 5 | $0.00024 | $0.00024 |
| Haiku 4.5 | $0.00012 | $0.00012 |
Grade A, and why
playwright copilot-instructions.md 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 yesterday.
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.
What it actually says
PR Review Guidelines
When reviewing pull requests:
- Only comment on semantically meaningful issues: bugs, incorrect logic, security problems, or API contract violations.
- Skip style, formatting, naming, and whitespace observations unless they cause functional problems.
- Keep each comment short — one or two sentences maximum.
- Do not write long descriptions or summaries of what the code does.
- Do not suggest refactors or improvements unrelated to the PR's stated goal.
- NEVER produce a review body or top-level PR overview comment. The review body MUST be empty. Put findings only in inline comments on specific lines.
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.
- yesterday First seen · 11 lines · 121 tokens per session scan A d28cb21c9dcb
playwright copilot-instructions.md is an instructions file published in the GitHub repository microsoft/playwright (95,378 stars, last pushed 3d ago), licensed Apache-2.0. It adds 121 tokens to every session, about $0.0006 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 instructions, from other repositories
test-proxy-recorder CLAUDE.md
Instructions for asmyshlyaev177/test-proxy-recorder: See AGENTS.md for the repo structure cheatsheet, commands, and conventions.
test-proxy-recorder AGENTS.md
Instructions for asmyshlyaev177/test-proxy-recorder, covering agents — structure cheatsheet, layout, commands (run from repo root), site audits (docs site) and the accessibility gate.
claude-browser-test-skills CLAUDE.md
Instructions for Autodesk/claude-browser-test-skills, covering claude.md, what this repo is, layout, conventions when editing skills and validate before publishing.
playwright-fieldkit AGENTS.md
Instructions for jpbaking/playwright-fieldkit: Follow the Portable Agent Authoring guides for every skill or harness change.
AI-Watch-Tester CLAUDE.md
Instructions for ksgisang/AI-Watch-Tester, covering aat (ai auto tester) — project guide, ⛔ awt testing workflow (필수 준수), step 1: scan, step 2: generate + present and step 3: execute (승인 후에만).
AI-Watch-Tester AGENTS.md
Instructions for ksgisang/AI-Watch-Tester, covering awt (ai watch tester) — ai agent rules, ⛔ mandatory workflow — 4 steps, no exceptions, step 1: scan, step 2: generate + present and step 3: execute (only after user says "go ahead").