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 instructions/mstocker1/visual_annotation_mcp/claude-mdgit clone --depth 1 https://github.com/mstocker1/Visual_Annotation_MCPWrote 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/instructions/mstocker1/visual_annotation_mcp/claude-md)<a href="https://agentmods.dev/instructions/mstocker1/visual_annotation_mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/mstocker1/visual_annotation_mcp/claude-md.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 | $0.00450 | $0.00450 |
| Opus 5 | $0.00225 | $0.00225 |
| Sonnet 5 | $0.00090 | $0.00090 |
| Haiku 4.5 | $0.00045 | $0.00045 |
Grade A, and why
Visual_Annotation_MCP CLAUDE.md 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 3d 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.
What it actually says
Visual Annotation MCP — Claude Code
One-time setup (this repo)
-
Create a venv and install (from repo root):
python -m venv .venvWindows:
.venv\Scripts\pip install -e .then.venv\Scripts\python -m playwright install chromiummacOS/Linux:
source .venv/bin/activate && pip install -e . && python -m playwright install chromium -
MCP config is in
.mcp.json(project scope). Claude Code may ask you to approve project MCP servers once; if tools do not appear, runclaude mcp reset-project-choicesand restart. -
If the server fails to start because
pythonis not on PATH for the GUI (common on Windows), either:- set
VISUAL_ANNOTATION_PYTHONto your venv interpreter before starting Claude Code (full path topython.exeorpython), or - replace the
commandin.mcp.jsonwith that full path, or - on Windows only, point
command/argsatpowershell+-File scripts/run_visual_mcp.ps1(run from this repo root).
- set
-
First browser launch can be slow; if needed start Claude with a higher MCP startup timeout, e.g.
MCP_TIMEOUT=120000(units per your shell / Claude docs). -
In Claude Code, run
/mcpand confirm visual-annotation is connected. Tools:navigate,inspect_elements,screenshot_viewport,screenshot_element,highlight_element,annotate_last_image.
Quick manual test prompt
After opening a project that contains this .mcp.json, ask Claude to: navigate to https://example.com, run inspect_elements, take a viewport screenshot, and describe what it sees.
Optional host allowlist: set VISUAL_ANNOTATION_ALLOWED_HOSTS (comma-separated hostnames, no https://) in the server env block of .mcp.json if you want to restrict URLs.
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.
- 3d ago First seen · 29 lines · 450 tokens per session scan A 83cd7d81869f
Visual_Annotation_MCP CLAUDE.md is an instructions file published in the GitHub repository mstocker1/Visual_Annotation_MCP (0 stars, last pushed 4mo ago), licensed MIT. It adds 450 tokens to every session, about $0.0022 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 instructions, from other repositories
Design-Learn CLAUDE.md
Instructions for GalaxyXieyu/Design-Learn, covering claude.md, todo, quick start, server and vscode extension.
landingforge CLAUDE.md
Claude Code instructions for NmadeleiDev/landingforge, covering landingforge — project guide for agents, browser automation (required for workflows and agents), conventions and skill routing.
AIPex CLAUDE.md
Instructions for AIPexStudio/AIPex, a project described as: AIPex: AI browser automation assistant, no migration and privacy first. Alternative to Manus Browser Operator、 Claude Chrome and Agent Browser.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.