AInotate

Annotated screenshots for AI agents: the agent shows you the button instead of describing where it is.

Ask an agent how to post a link on Hacker News and you usually get directions: the orange bar at the top, the last link after jobs. With AInotate it sends the picture above instead and says "click submit, where the blue arrow points".

It captures a web page, a window or a screenshot you already have, finds the elements you mean, draws numbered steps, arrows, boxes and labels where they hide the least, blacks out emails and tokens, and looks at the result before sending it. It works in Claude Code, Cowork, Claude Desktop, Cursor and any other MCP client, and from the command line or Python.

We built it at DontPayFull because our own agents kept answering "where is it?" with a paragraph. A picture with a number on the right button settles the question in a second, in a chat, a ticket or a guide.

What it is for

The obvious case is "where is it?": a setting three menus deep, a button that is only an icon. The agent replies with the screen, the control boxed and numbered. For "walk me through it" it makes one image per step, and ainotate guide binds them into a Markdown, HTML or PDF guide.

Some steps belong to you, not the agent: a permission switch, a 2FA code, a payment or consent screen. The agent doesn't click those. It shows what to press and says what the click will do.

It also helps with bug reports (what is wrong in red, what is right in green, personal data blacked out before anyone sees it) and with docs, where the screenshot points at the thing the text is about and can be re-shot from the same spec when the UI changes.

Set it up in your agent

1. Install, one of:

Then ainotate doctor checks the machine and prints the exact fix for anything missing, including the one command that downloads Chromium for web capture. To update later: brew upgrade ainotate, pipx upgrade ainotate or uv tool upgrade ainotate; the plugin with /plugin marketplace update ainotate; the Desktop extension by opening the newer .mcpb from Releases.

2. Connect it to the agent you use:

{"mcpServers": {"ainotate": {"command": "ainotate", "args": ["mcp"]}}}

AInotate runs on your own computer: Cowork reaches it through the Claude desktop app, so keep the app open while a task uses it. The plugin starts the installed ainotate, or runs it with uvx when it is not installed.

The skill teaches an agent when and how to annotate: pick a source, find exact positions, annotate, read the image to verify, deliver. The plugin brings it along; without the plugin it ships with the package:

ainotate install-skill             # Claude Code and Codex (~/.claude/skills, ~/.agents/skills)
ainotate install-skill --zip ~/Desktop   # a ZIP for Claude Desktop, Cowork, claude.ai:
                                         # Customize > Skills > Upload a skill

Config paths for every OS, permissions and the tool list: skill/references/mcp.md.

3. Ask. "Show me where to switch Wikipedia to dark mode." The agent captures the page, annotates it, checks it and sends the image. With the skill it also does this on its own whenever an answer depends on where something is on the screen: a "how do I" question, a visual bug, a before/after of a change.

Gallery

Every image below was made in one shoot call against a public site: open the page, find the elements, place the labels, draw, frame. The specs are in docs/images/specs.

Numbered steps on the busy header of DontPayFull, coupons for 20,000+ stores	Box, arrow and highlight on GitHub
Step-by-step guide with browser chrome and a gradient frame	Bug report, emails redacted automatically by the privacy scan
Magnifier and keycaps for tiny controls and shortcuts	Before / after plate from two shots

Four arrow styles: skitch (default: straight when the way is clear, a gentle bend when it is long and diagonal, a curve around text), curved, straight, line

Phone preset: mobile viewport, touch and user agent

Labels that stay off the content

A label dropped at a fixed spot next to its target usually lands on the text the reader needs. AInotate places every label itself:

• It maps the page content first (text, icons, lines, images) and tries hundreds of spots per label. Empty space near the target wins; a spot over text is used only when the frame has no free room.
• Arrows are priced too: a long arrow, or one that crosses text, another label or another mark, loses to a shorter, cleaner one. When the straight path crosses text, the arrow curves around it.
• All labels are planned together, so the first one cannot take the only clean spot a later one needs.
• Step badges sit on the corner of their box that covers the least.

The image is never enlarged to make room. If a label still has to cover something, AInotate says so in a warning, and label_at pins a label exactly where you want it.

From the command line

Save this as hn.json:

{"url": "https://news.ycombinator.com",
 "frame": {"chrome": "browser"},
 "marks": [
  {"type": "step", "n": 1, "target": {"text": "new", "exact": true},
   "label": "Newest"},
  {"type": "step", "n": 2, "target": {"text": "login"},
   "label": "Sign in"}]}

ainotate shoot hn.json --draft    # temp file; drop --draft to save

The output path is printed on stdout. Images are saved to ~/Pictures/AInotate unless you configure another folder. ainotate --help lists every command (capture, annotate, locate by OCR, grid, zoom, window capture with UI element targets, elements, guide, compare, animate, copy, install-skill); the full spec is in skill/references/spec.md. Exit codes: 0 ok, 1 cannot save, 2 invalid spec, 3 cannot draw, 4 capture failed or an optional dependency is missing, 5 ambiguous text target, 6 target not found.

Python. The same pipeline is importable:

from ainotate.shoot import shoot
res = shoot({"url": "https://en.wikipedia.org/wiki/Screenshot",
             "marks": [{"type": "box",
                        "target": {"text": "View history"},
                        "label": "Past edits"}]}, draft=True)
print(res.paths[0], res.redactions)

Features

What it draws. Numbered steps, boxes, arrows with labels, a click ripple, keycaps for shortcuts, a magnifier for small controls, highlights, a spotlight that dims everything else, free text, and solid redaction (blur and pixelate exist too, for clutter only). Labels go to free space and are planned together, as described above. It warns when labels overlap, arrows cross, or a frame carries more than six marks.

Where the images come from. ainotate shoot opens a page in a laptop, wide or phone viewport, runs clicks and form fills, finds each target by text, role or CSS selector, redacts, draws and saves in one call. It uses its own Chromium, Firefox or WebKit, or attaches over CDP to a Chromium you are already logged into (Chrome, Edge, Brave, Arc, BrowserOS) and puts that tab back the way it was. Any other image works too: locate finds text by OCR (Apple Vision, Windows OCR or Tesseract), grid and zoom pin down icons, and screen, window and clipboard capture are built in. On macOS, ainotate window ID --target reads buttons, switches and fields straight from the Accessibility API, so a mark sits exactly on the control even when it is just an icon.

What you can do with them. Frame them on a gradient with browser or window chrome, build guides in Markdown, HTML or PDF, put a before and an after side by side, make an APNG or GIF, or copy the result to the clipboard. A spec with a typo stops with every problem listed rather than saving a wrong or unredacted image, and every failure has its own exit code. The bundled Inter font renders accents and diacritics the same on every machine.

Privacy

Redaction paints a solid block. AInotate never blurs data to hide it, because blur and pixelation can be reversed; marks flagged "sensitive": true are refused for them.

On web shots automatic redaction is on: before the screenshot, shoot reads the live page, input values included, and blacks out emails, phone numbers, card numbers, IBANs, API keys, tokens, JWTs and password or personal-data fields. On images it does the same by OCR when you pass --privacy auto. You can allow-list your own addresses and add patterns.

Detection is best-effort. It can miss text drawn into images or canvas, odd formats, data split across elements, and frames embedded from other sites (those are reported, not redacted). Look at the image before you share it; the skill makes agents do exactly that. Everything runs on your machine, and AInotate makes no network requests beyond the pages you ask it to capture.

Platform support

Experimental means implemented and unit-tested with mocks, not yet verified on real Windows or Linux machines. Reports are welcome.

License

Copyright © 2026 DontPayFull. AInotate is free software under the GNU Affero General Public License v3.0 or later. If you run a modified AInotate as a network service, the AGPL requires you to offer its source to the users of that service.

Bundled and derived third-party work (the Inter font, arrow proportions from Arrowshot, a label-placement idea from github/awesome-copilot) and dependency licenses are listed in NOTICE.

Made with ❤️ by the DontPayFull team
_{Coupons & discount codes for 20,000+ stores}