browser-testing

A browser-based test guide for checking a floating-point conversion calculator. It covers visual checks, button interactions, and whether displayed conversions match known values.

In plain words
What is it for?
Use it to open the calculator in a browser, take screenshots, test preset buttons and format conversions, and check that binary and hexadecimal displays update correctly.
Why use it?
It helps catch broken layouts, unresponsive controls, incorrect conversions, and changes that damage the web interface.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/sw23/fp-conv/browser-testing
Any agent
npx skills add sw23/fp-conv --skill browser-testing
Clone the repo
git clone --depth 1 https://github.com/sw23/fp-conv

Made for: Claude Code, Codex.

Per session 60 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,254 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00060 $0.02254
Opus 5 $0.00030 $0.01127
Sonnet 5 $0.00012 $0.00451
Haiku 4.5 $0.00006 $0.00225

Measured 2d ago against content hash ba907dded3a2, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

browser-testing scanned grade B with 1 finding 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.

Unrestricted tool accessmediumExcessive agency

A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.

2. **Enable browser tools in chat**: Open the Tools picker in the chat input and enable all tools under Built-in > Browser
.github/skills/browser-testing/SKILL.md · 189 lines

How it starts

The opening of the file, as written. The whole thing — 189 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Browser UI Testing

Test the floating-point conversion calculator in an actual browser using VS Code's browser agent tools. Supports three testing modes: visual verification, interaction testing, and conversion validation.

Prerequisites

  1. Enable browser tools: Set workbench.browser.enableChatTools to true in VS Code settings
  2. Enable browser tools in chat: Open the Tools picker in the chat input and enable all tools under Built-in > Browser
  3. Use #tool:openBrowserPage to open index.html from the workspace root as a file:// URL (no local server needed — the built-in browser loads local files and scripts correctly). Construct the URL from the workspace path, e.g. file://${workspaceFolder}/index.html

Reference Files

  • UI element map — All DOM selectors, element IDs, button data attributes
  • Test vectors — Known-good encode/decode/conversion values for validation

Procedure A — Visual Verification

Verify the page renders correctly and all sections are present.

Steps

  1. Open the page with #tool:openBrowserPage
  2. Take an initial screenshot with #tool:screenshotPage — verify:
    • Header displays "Floating Point Conversion Calculator"
    • Subtitle "Convert between floating point, integer, and custom formats" is visible
    • All four main sections are visible: Input Format, Input Value, Output Format, Output Value
    • Format preset buttons are rendered in groups (IEEE 754, ML, OCP, Integer)
    • Binary bit display shows clickable checkboxes
    • About section is present at the bottom
  3. Use #tool:readPage to confirm page text content includes:
    • All format button labels: FP64, FP32, FP16, BF16, TF32, FP8 E5M2, FP8 E4M3, FP6 E3M2, FP6 E2M3, FP4 E2M1
    • All integer format labels: INT32, UINT32, INT16, UINT16, INT8, UINT8, INT4, UINT4
    • Value preset labels: 0, 1, Max Norm, Min Norm, Max Sub, Min Sub, +Inf, -Inf, NaN, 1s
    • Rounding mode options: Nearest Ties to Even, Ties Away, Toward Zero, Toward +Infinity, Toward -Infinity
  4. Verify the default initial state matches:
    • Input format: FP16 (total bits = 16)
    • Output format: BF16 (total bits = 16)
    • Decimal value: 3.140625

Read the full file on GitHub · 189 lines

Files

What ships with it

2 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.

Changes

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.

  1. 2d ago First seen · 189 lines · 60 tokens per session scan B ba907dded3a2

Subscribe to this mod's changes

browser-testing is a skill published in the GitHub repository sw23/fp-conv (5 stars, last pushed 6d ago), licensed MIT. It adds 60 tokens to every session and 2,254 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it B with 1 finding (unrestricted tool access). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories