Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add ericboehs/claude-plugins/plugin install agent-desktopWrote 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/skills/ericboehs/claude-plugins/agent-desktop)<a href="https://agentmods.dev/skills/ericboehs/claude-plugins/agent-desktop"><img src="https://agentmods.dev/badge/skills/ericboehs/claude-plugins/agent-desktop/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/skills/ericboehs/claude-plugins/agent-desktop"><img src="https://agentmods.dev/badge/skills/ericboehs/claude-plugins/agent-desktop.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 | $0.00068 | $0.02514 |
| Opus 5.5 | $0.00027 | $0.01006 |
| Sonnet 5.5 | $0.00014 | $0.00503 |
| Haiku 4.5 | $0.00007 | $0.00251 |
Grade A, and why
agent-desktop 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 2mo 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 — 134 lines — stays where its author put it; the contents beside it link to each section on GitHub.
agent-desktop (macOS accessibility automation)
Drive native macOS apps through the OS accessibility tree with the agent-desktop CLI. It is NOT an agent; it is a tool you (the agent) invoke. Every command prints a JSON envelope on stdout; you run the observe→act→observe loop.
Prerequisites
agent-desktop must be on PATH and the launching terminal needs Accessibility permission (Screen Recording too, for screenshots). Check with agent-desktop status; it reports both permissions and the latest snapshot. If the command is missing or a binary error appears, suggest /setup-agent-desktop.
The loop
snapshot → decide → act → snapshot → decide → act → ...
Refs (@e1, @e2, …) are assigned per snapshot and scoped to a snapshot_id (e.g. s8f3k2p9). Pass --snapshot <id> on ref commands for determinism; it does not require --session.
Dense apps: skeleton then drill (saves 60–96% tokens)
# 1. shallow depth-3 map; truncated containers show children_count and get refs
agent-desktop snapshot --skeleton --app Xcode -i --compact # keep snapshot_id
# 2. drill into ONE region instead of re-dumping the whole tree
agent-desktop snapshot --root @e11 --snapshot <id> -i --compact
# 3. act on something revealed in the drill
agent-desktop click @e57 --snapshot <id>
# 4. re-drill the same region to verify (scoped invalidation: only @e11's subtree refs change)
agent-desktop snapshot --root @e11 --snapshot <id> -i --compact
Simple apps: full snapshot is fine
agent-desktop snapshot --app Finder -i --compact # or use `find` if you know the element
agent-desktop find --role button --app TextEdit
Electron apps: enable accessibility first, THEN snapshot
Electron/Chromium apps (Slack, VS Code, Discord, Notion, Obsidian, Teams, Figma) ship their full a11y tree turned OFF for performance. Until something flips it on, every snapshot returns only a ~3-ref stub. Set AXManualAccessibility=YES on the app once per launch with the bundled helper, then proceed as normal. On a live test this took Slack from 3 refs to 534.
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.
- 2mo ago First seen · 134 lines · 68 tokens per session scan A 8d54a23731ff
agent-desktop is a skill published in the GitHub repository ericboehs/claude-plugins (2 stars, last pushed 2mo ago), licensed MIT. It adds 68 tokens to every session and 2,514 once invoked, about $0.0003 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.
agent-desktop compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| agent-desktopericboehs/claude-plugins | Skill | 2 | 2mo ago | MIT | A | 68 | Claude Code |
| imagegenopenai/codex | Skill | 134k | today | Apache-2.0 | A | 113 | Codex |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 79k | today | MIT | A | 200 | Claude Code, Codex |
| playwright-climicrosoft/playwright | Skill | 99k | today | Apache-2.0 | A | 19 | Claude Code |
| systematic-debuggingobra/superpowers | Skill | 314k | 4d ago | MIT | A | 21 | Claude Code |
| next-cache-components-adoptionvercel/next.js | Skill | 143k | today | MIT | A | 95 | Claude Code |
| chat-pet-sprite-creationmicrosoft/vscode | Skill | 195k | today | MIT | A | 53 | Claude Code, Codex |