verify

verify is a skill for Claude Code from anthropics/claude-quickstarts. It costs 23 tokens per session (1,597 once invoked), scanned A, original, MIT.

A verification guide for the computer-use demo's Streamlit web app. Streamlit is a tool for building browser-based Python interfaces, and the guide tests the finished app inside a Docker container with Playwright, a browser-automation tool.

In plain words
What is it for?
It helps build and run the Docker image, wait for the app to start, replace source files during iteration, and drive the interface with automated browser tests.
Why use it?
It checks the actual packaged application end to end, so passing code checks alone is not mistaken for a working user interface.

Skill for Claude Code

Written for Claude Code: installed under .claude/.

About the project

Claude Quickstarts is a collection of starter projects for building applications with the Claude API, including customer support, financial analysis, and computer-use examples. Developers use them as foundations to customize for their own deployable applications, and the catalogue add-ons provide workflows for working with these quickstarts.

anthropics/claude-quickstarts · 17,608 stars · on GitHub

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/anthropics/claude-quickstarts/verify
Any agent
npx skills add anthropics/claude-quickstarts --skill verify
Clone the repo
git clone --depth 1 https://github.com/anthropics/claude-quickstarts

Made for: Claude Code.

Wrote 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.

agentmods badge for verify

README.md
[![agentmods](https://agentmods.dev/badge/skills/anthropics/claude-quickstarts/verify.svg)](https://agentmods.dev/skills/anthropics/claude-quickstarts/verify)
Your own site
<a href="https://agentmods.dev/skills/anthropics/claude-quickstarts/verify"><img src="https://agentmods.dev/badge/skills/anthropics/claude-quickstarts/verify.svg" alt="Measured on agentmods" height="20"></a>
Per session 23 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,597 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 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 5.1 $0.00023 $0.01597
Opus 5 $0.00012 $0.00798
Sonnet 5 $0.00005 $0.00319
Haiku 4.5 $0.00002 $0.00160

Measured 6d ago against content hash 6b03eb806887, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

verify scanned grade A 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 6d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

until curl -sf localhost:18501/_stcore/health >/dev/null; do sleep 1; done
computer-use-demo/.claude/skills/verify/SKILL.md · 137 lines

How it starts

The opening of the file, as written. The whole thing — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Verifying changes to this demo

The surface is the Streamlit app inside the container. Build the image from this directory, run it, and drive the UI with Playwright. pytest, ruff, and pyright are CI's job, not evidence.

Build and run

docker build . -t computer-use-demo:verify
docker rm -f cu-verify 2>/dev/null
docker run -d --name cu-verify \
  -e ANTHROPIC_API_KEY=$ANTHROPIC_API_KEY -e WIDTH=1024 -e HEIGHT=768 \
  -p 18501:8501 -p 18080:8080 -p 16080:6080 computer-use-demo:verify
until curl -sf localhost:18501/_stcore/health >/dev/null; do sleep 1; done

A cold build takes several minutes (apt, LibreOffice, a pyenv Python compile). The requirements.txt layer is cached separately, so source-only changes rebuild in seconds.

To iterate without rebuilding, copy sources over the baked copy before the first browser session compiles the script:

docker cp computer_use_demo/. cu-verify:/home/computeruse/computer_use_demo/

Streamlit's file watcher does not notice docker cp replacements once a session has run. Recreate the container instead of docker restart, which fails on a stale X lock (tint2: could not open display).

Drive the UI

Install Playwright in any venv (pip install playwright && python -m playwright install chromium) and run the script below. It sets the model, picks a tool version, optionally sets the thinking mode, sends a prompt, waits for the loop to go idle, then dumps the chat transcript and every HTTP exchange (request headers and JSON, response JSON) with auth headers redacted.

THINKING=Adaptive python drive.py 18501 claude-sonnet-5 computer_toolset_20260801 \
  "Take a screenshot, then zoom in on the taskbar and list the icons." out/ 240

Pass - for model or tool version to keep the default. THINKING=Off|Adaptive|Extended clicks that radio. TOKEN_EFFICIENT=1 ticks the beta checkbox.

What to read afterwards:

  • out/chat.txt: the rendered conversation, including Tool Use: <name> lines and any error box with its traceback.
  • out/exchange-NN.txt: one file per API round trip. Check tools, anthropic-beta, and the tool_use / tool_result blocks here rather than trusting the chat rendering.
  • out/02-chat.png, out/04-http-logs-expanded.png: full-page screenshots.

Read the full file on GitHub · 137 lines

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. 6d ago First seen · 137 lines · 23 tokens per session scan A 6b03eb806887

Subscribe to this mod's changes

verify is a skill published in the GitHub repository anthropics/claude-quickstarts (17,608 stars, last pushed yesterday), licensed MIT. It adds 23 tokens to every session and 1,597 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.