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/magus --skill browser-debugginggit clone --depth 1 https://github.com/MadAppGang/magusWrote 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/magus/browser-debugging)<a href="https://agentmods.dev/skills/madappgang/magus/browser-debugging"><img src="https://agentmods.dev/badge/skills/madappgang/magus/browser-debugging/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/magus/browser-debugging"><img src="https://agentmods.dev/badge/skills/madappgang/magus/browser-debugging.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 6 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium MCP Rug Pull · line 109 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 176 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 305 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 370 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 421 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium Memory Poisoning · line 523 Skill attempts to fill the context window with filler content, displacing legitimate instructions and safety constraints. This can degrade agent performance or bypass safety boundaries.Fix: Implement context-window management that detects and rejects padding or stuffing attempts. Prioritize system instructions over user-injected content.
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 | $0.00051 | $0.02204 |
| Opus 5.5 | $0.00020 | $0.00882 |
| Sonnet 5.5 | $0.00010 | $0.00441 |
| Haiku 4.5 | $0.00005 | $0.00220 |
Grade A, and why
browser-debugging 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 22d 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 — 175 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Browser debugging
Settle a UI question with evidence from a running browser — what renders, what the console reports, what the network did — instead of inferring it from source. A finding is something you observed on the page. Say which revision of the code it depicts, at which URL and viewport, and what you did not check.
Paths in this file are relative to the dev plugin root: the directory that holds
skills/ and knowledge/, three levels above this file's folder.
Who can run this
Browser tools live in the main session. No dev agent has them — every agent in agents/
declares a tools: list of file and shell tools only. An agent that reads this file
without browser tools reports the browser check as not performed and names the
observation that would settle it. It never describes a page it did not see.
When a subagent needs visual evidence, the main session supplies it as a file path;
dev:frontend-developer reads images with Read. Producing that file is not automatic —
see Screenshots as files.
Choose the browser
Two MCP servers can provide a browser, and either may be absent. Their tools are often deferred — listed by name, with the schema loaded through ToolSearch — so look for the prefix in the tool list rather than probing with a call.
| claude-in-chrome | browser-use@magus | |
|---|---|---|
| Tool prefix | mcp__claude-in-chrome__ |
mcp__plugin_browser-use_browser-use__ |
| Browser | the user's own Chrome, with their logins | a Chromium the server launches, fresh profile, headed by default |
| Console | read_console_messages |
no tool |
| Network | read_network_requests |
no tool |
| Screenshot | computer screenshot action; zoom for a region |
browser_screenshot, optionally full page |
| Viewport size | resize_window |
no tool |
| JavaScript in the page | javascript_tool |
browser_evaluate |
| Addressing elements | ref_N from read_page or find, or coordinates |
index from browser_get_state, or coordinates |
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.
- 22d ago Changed · -481 lines · +17 tokens per session 49bf49430a01
- 1mo ago First seen · 656 lines · 34 tokens per session scan A bde696cfe26c
browser-debugging is a skill published in the GitHub repository MadAppGang/magus (13 stars, last pushed today), licensed MIT. It adds 51 tokens to every session and 2,204 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-04.
Other skills, from other repositories
chrome-devtools-cli
Use this skill to write shell scripts or run shell commands to automate tasks in the browser or otherwise use Chrome DevTools via CLI.
a11y-debugging
Uses Chrome DevTools MCP for accessibility (a11y) debugging and auditing based on web.dev guidelines. Use when testing semantic HTML, ARIA labels, focus states, keyboard navigation, tap targets, and color contrast.
chrome-devtools
Uses Chrome DevTools via MCP for efficient debugging, troubleshooting and browser automation. Use when debugging web pages, automating browser interactions, analyzing performance, or inspecting network requests. This skill does not apply to --slim mode (MCP configuration).
opencli-autofix
Automatically fix broken OpenCLI adapters when commands fail. Load this skill when an opencli command fails — it guides you through collecting a trace artifact, patching the adapter, retrying, and filing an upstream GitHub issue after a verified fix. Works with any AI agent.
chrome-performance-optimizer
Autonomous multi-agent performance optimization loop for Chromium and V8. Supports profile-seeded mode (analyzing Speedometer 3 / JetStream profiles via pprof or Sagacity MCP) and pattern-driven discovery mode (fan-out exploration of Blink and V8 macro-patterns grounded in historical wins and past rejected CLs).…
nullaway
Guide for resolving NullAway static analysis errors. Best practices for: Passing ObservableSupplier/Supplier Dereferencing potentially @Nullable values Adding @NullMarked to Java code.