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 Razaib-khan/ForgeWeave --skill playwright-mcpgit clone --depth 1 https://github.com/Razaib-khan/ForgeWeaveWrote 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/razaib-khan/forgeweave/playwright-mcp)<a href="https://agentmods.dev/skills/razaib-khan/forgeweave/playwright-mcp"><img src="https://agentmods.dev/badge/skills/razaib-khan/forgeweave/playwright-mcp/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/razaib-khan/forgeweave/playwright-mcp"><img src="https://agentmods.dev/badge/skills/razaib-khan/forgeweave/playwright-mcp.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 | $0.00156 | $0.01528 |
| Opus 5 | $0.00078 | $0.00764 |
| Sonnet 5 | $0.00031 | $0.00306 |
| Haiku 4.5 | $0.00016 | $0.00153 |
Grade A, and why
playwright-mcp 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 8d 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 — 184 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Playwright MCP — AI Agent Browser Automation
Playwright MCP is a Model Context Protocol server that lets AI agents control real browsers through structured accessibility snapshots. Refs like e5, e12 give deterministic handles to every interactive element — no pixel-guessing, no brittle CSS selectors.
Read
references/tools.mdfor the full tool catalogue before choosing an approach.
Readreferences/config.mdfor CLI flags, capabilities, and JSON config options.
Readreferences/patterns.mdfor copy-paste workflow recipes.
Core Mental Model
browser_navigate → browser_snapshot → interact via refs → browser_snapshot (verify)
Every action follows this cycle. The snapshot is the ground truth. Never assume page state — always re-snapshot after an action.
Decision Tree: What to Use
1. Snapshot vs Vision Mode
| Situation | Use |
|---|---|
| Standard web app — forms, buttons, links, tables | Snapshot (default) |
| Canvas, SVG, chart, game UI, custom drawn widget | Vision mode (--caps=vision) |
| Verifying visual layout, pixel correctness | Screenshot + snapshot together |
Rule: Snapshots for 90%+ of tasks. Vision only for elements absent from the accessibility tree.
2. Which Capability to Enable
Enable only what you need — each capability adds tokens to every interaction:
| Need | Add to --caps= |
|---|---|
| Mock API responses / intercept network | network |
| Cookies, localStorage, auth state save/restore | storage |
Verify elements, generate .spec.ts test code |
testing |
| Canvas/chart coordinate-based clicking | vision |
| Export page as PDF | pdf |
| Video recording, tracing | devtools |
Minimal config (no capabilities) is fastest and cheapest on tokens.
3. Token Budget Awareness
A single browser_navigate on a content-rich page can return thousands of tokens. Across a multi-step session this compounds.
| Practice | Saves Tokens |
|---|---|
| Enable only needed capabilities | High |
Use browser_snapshot only when you need to verify or find a ref |
High |
Avoid browser_take_screenshot unless visual proof is required |
Medium |
Use browser_fill_form (one call) instead of multiple browser_type |
Low |
| Stop tracing/video when not debugging | Medium |
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 8d ago First seen · 184 lines · 156 tokens per session scan A cb35d94cefe9
playwright-mcp is a skill published in the GitHub repository Razaib-khan/ForgeWeave (1 stars, last pushed 2mo ago), licensed MIT. It adds 156 tokens to every session and 1,528 once invoked, about $0.0008 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 2026-08-31.
Other skills, from other repositories
browser-automation
A browser-automation workflow that detects the terminal environment and chooses a browser control tool. It supports opening pages, taking snapshots, clicking, filling fields, and other checks.
playwright-cli
Skill "playwright-cli" from Insajin/autopus-adk, covering browser automation with playwright-cli, quick start, commands, core and navigation.
structured-extraction
Extract structured data from web pages using browser snapshot and text tools, then process it into tables, comparisons, or summaries using Python.
web-researcher
Web research tools for fetching and reading web pages. Use when the agent needs to browse the web or make HTTP API calls.
runtime-surface-probing
Plan and interpret safe runtime checks against a locally running application.
computer-use
OS/window-level inspection and input in visible local app windows through orca computer: native apps, external browser windows (Chrome, Edge, Safari), and app webviews. Not for Orca's embedded browser (use orca-cli) or page-only automation (use Playwright or CDP).