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 skills/managedcode/dotnet-skills/chrome-devtools-mcpnpx skills add managedcode/dotnet-skills --skill chrome-devtools-mcpgit clone --depth 1 https://github.com/managedcode/dotnet-skillsWrote 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/managedcode/dotnet-skills/chrome-devtools-mcp)<a href="https://agentmods.dev/skills/managedcode/dotnet-skills/chrome-devtools-mcp"><img src="https://agentmods.dev/badge/skills/managedcode/dotnet-skills/chrome-devtools-mcp.svg" alt="Measured on agentmods" 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 | $0.00130 | $0.02033 |
| Opus 5 | $0.00065 | $0.01017 |
| Sonnet 5 | $0.00026 | $0.00407 |
| Haiku 4.5 | $0.00013 | $0.00203 |
Grade A, and why
chrome-devtools-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 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.
How it starts
The opening of the file, as written. The whole thing — 173 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Chrome DevTools MCP for .NET Repos
Trigger On
- the repo needs browser-level debugging for ASP.NET Core, Blazor, WebAssembly, or any .NET app with a web UI
- the user wants an MCP server that can inspect console output, network traffic, screenshots, traces, and DOM state
- the repo needs agent-friendly browser control instead of manual DevTools work
Do Not Use For
- pure .NET code analysis, unit testing, or NuGet/package management
- static HTML linting alone
- sensitive browser sessions that must not be exposed to an MCP client
Inputs
- the nearest
AGENTS.md - the MCP client config for the current agent or app
- the target Chrome mode: launch new, connect to existing browser, or attach to a WebSocket endpoint
- the target app URL or local dev server
Workflow
- Choose the browser mode first.
- new ephemeral browser: default when you want a clean session
- existing browser: use
--browserUrlor--wsEndpoint - running browser with a user profile: use
--autoConnectonly when Chrome 144+ and remote debugging are already enabled
- Prefer repo-local config over ad hoc shell commands.
- store the MCP server block in the repo or agent config so every run is repeatable
- Start from the smallest useful surface.
- use
--slimwhen you only need navigation, JavaScript execution, and screenshots - leave the full toolset enabled when you need network, performance, or emulation
- use
- Be explicit about privacy and cost.
- disable usage statistics or CrUX-backed performance enrichment when policy or sensitivity requires it
- Validate against the real app, not a mock page.
- use the app's local dev server, preview URL, or a production-like environment that reflects the bug
Current Upstream Notes
- Chrome DevTools MCP
v1.8.0adds PWA automation,query_heapsnapshot, richer heap-edge and native-context inspection, console stack traces, multi-file upload, and a stable-DOM bypass forevaluate_script. Enable memory-debugging tools explicitly and keep heap output bounded to the investigation. - Page-scoped tools now require
pageIdby default. Update callers to select and pass the intended page instead of relying on ambient tab state. The release also fixes CLI default handling, remote-path validation, HiDPI screenshot bounds, slow-client root listing, and lifecycle regressions; reproduce the real browser path before adding client-side workarounds. - Use the memory tools only when heap snapshots are the actual debugging need; close snapshots when finished so the browser session does not keep unnecessary memory pressure.
- For sensitive or scoped browser automation, prefer
allowedUrlPattern/blockedUrlPatternover relying only on prompt instructions. - When enabling category extensions, prompt the user for the category instead of silently broadening the tool surface.
What ships with it
1 file 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.
- yesterday First seen · 173 lines · 130 tokens per session scan A 140d6f652ec1
chrome-devtools-mcp is a skill published in the GitHub repository managedcode/dotnet-skills (477 stars, last pushed 4d ago), licensed MIT. It adds 130 tokens to every session and 2,033 once invoked, 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 2026-09-03.
Other skills, from other repositories
scrapling
使用 scrapling 进行网页抓取和数据提取。自动选择 Fetcher，支持 Cloudflare/WAF 绕过、Session 登录、HTML 解析。当用户提到 scrape/crawl/fetch page/extract data/爬取/抓取/绕过Cloudflare/解析HTML/批量采集 时触发。.
close-task-commit-push-pr
Close the active backlog task (detected from branch name), commit all changes, push to remote, and open a pull request. Use when the user says "close task and ship it", "close task commit push pr", or invokes /close-task-commit-push-pr.
macos-use
GUI control for macOS apps via mediar-ai's mcp-server-macos-use. Click, type, scroll, key-press, open apps — driven by accessibility tree, works in non-interactive Claude Code mode. Use this for any Sutando task that needs to drive another macOS application (Safari, Zoom, Mail, Finder, etc.).
cmux-browser
End-user browser automation with cmux. Use when you need to open sites, inspect or interact with browser surfaces, wait for page state, and extract data without stealing focus.
visual-verify-ui
Wraps the browser tool into a dedicated visual QA testing skill for frontend work.
mutation-testing
Use when running mutation testing, killing mutants, verifying test quality, checking mutation score, or analyzing survivors after the test baseline is green.