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.
git clone --depth 1 https://github.com/gettalon/talon-pluginsWrote 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/commands/gettalon/talon-plugins/computer)<a href="https://agentmods.dev/commands/gettalon/talon-plugins/computer"><img src="https://agentmods.dev/badge/commands/gettalon/talon-plugins/computer/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/commands/gettalon/talon-plugins/computer"><img src="https://agentmods.dev/badge/commands/gettalon/talon-plugins/computer.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00040 | $0.01842 |
| Opus 5.5 | $0.00016 | $0.00737 |
| Sonnet 5.5 | $0.00008 | $0.00368 |
| Haiku 4.5 | $0.00004 | $0.00184 |
Grade A, and why
computer 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 1mo 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 — 221 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Computer Use
Control the macOS desktop using the usecomputer CLI. 17 commands for mouse, keyboard, screenshots, windows, displays, clipboard, and coordinate mapping.
Setup
npm install -g usecomputer
Requirements: macOS, Accessibility permissions enabled for your terminal, Node.js 18+.
Core Workflow
- Screenshot → see what's on screen
- Analyze → find the target coordinates
- Act → click, type, or press keys (use
--coord-mapfrom screenshot) - Verify → screenshot again to confirm
Commands (17 total)
Screenshot
usecomputer screenshot /tmp/screen.png --json
usecomputer screenshot /tmp/region.png --region 0,0,800,600 --json
usecomputer screenshot /tmp/win.png --window 1234 --json
usecomputer screenshot /tmp/screen.png --display 1 --json
usecomputer screenshot /tmp/annotated.png --annotate --json
Options: --region x,y,w,h, --display N, --window ID, --annotate, --json
Returns JSON: path, coordMap, desktopIndex, imageWidth, imageHeight, hint
Always pass --coord-map from screenshot output when clicking/hovering/dragging.
Click
usecomputer click -x 500 -y 300 --coord-map "0,0,1792,1120,1568,980"
usecomputer click -x 500 -y 300 --button right --coord-map "..."
usecomputer click -x 500 -y 300 --count 2 --coord-map "..."
usecomputer click -x 500 -y 300 --modifiers ctrl,shift --coord-map "..."
Options: --button left|right|middle, --count N, --modifiers ctrl,shift,alt,meta, --coord-map
Hover — Move cursor without clicking
usecomputer hover -x 500 -y 300 --coord-map "..."
Mouse Move
usecomputer mouse move -x 500 -y 300 --coord-map "..."
Mouse Down / Up — Press and hold / release
usecomputer mouse down --button left
usecomputer mouse up --button left
Options: --button left|right|middle
Mouse Position
usecomputer mouse position --json
Type
usecomputer type "Hello world"
usecomputer type "slow text" --delay 20
cat file.txt | usecomputer type --stdin --chunk-size 4000 --chunk-delay 15
usecomputer type "limited" --max-length 1000
Options: --delay ms, --stdin, --chunk-size N, --chunk-delay ms, --max-length N
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.
- 1mo ago First seen · 221 lines · 40 tokens per session scan A 47972dddb6cd
computer is a command published in the GitHub repository gettalon/talon-plugins (6 stars, last pushed 6mo ago), licensed MIT. It adds 40 tokens to every session and 1,842 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 2026-08-31.
Other commands, from other repositories
qa
Dispatch a verifiable browser task to the qa-tester subagent. Use for regression checks, smoke tests, and any task with a clean pass/fail outcome. Supports an EXHAUSTIVE mode that verifies every control behind an honesty gate.
playbook
Manage Mochi playbooks (list, show, run, delete). Plays through the browser MCP tools.
unschedule-playbook
Cancel a scheduled playbook routine.
check-browser
Use Real Browser MCP to check what's currently showing in my browser.
checkpoint
Write a new link in the continuum chain summarizing decisions/changes since the last link.
design-qa
Agent-driven design QA — walk the running app, judge design/UX, and drop a Comment-Mode comment on every issue (visible to the human as pins). Then read the session back and fix everything in bulk. Optional focus argument steers the pass.