# lahfir/agent-desktop

Native desktop automation CLI for AI agents. Control any application through OS accessibility trees with structured JSON output and deterministic element refs.

Repository: https://github.com/lahfir/agent-desktop
Canonical: https://ross.abutalabs.com/products/agent-desktop
Language: Rust
License: Apache-2.0
License Family: permissive
Topics: accessibility, accessibility-api, ai-agents, automation, cli, desktop-automation, macos, mcp, rust
Last push: 2026-09-03T00:48:19+00:00

## Health v2 (maintenance only)
Score: 82/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 100, release rhythm 99, longevity 13
- inputs: {"age_days": 195, "days_push": 0, "days_rel": 6, "gap_med": 1, "n_releases_24m": 36}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1034, forks 70 (observed 2026-09-03T02:15:14.265387+00:00)

## What it is
A native Rust CLI that lets AI agents control desktop applications through OS accessibility trees, returning structured JSON with deterministic element references instead of screenshots. It also ships a C-ABI library for embedding from other languages and MCP integration.

## Use cases
- let an AI agent click buttons and fill forms in native desktop apps
- automate macOS applications without pixel-based screenshot matching
- give an LLM agent structured access to a GUI's accessibility tree
- reduce token usage when agents observe dense desktop UIs
- drive desktop apps from Python or Swift via a C-ABI library
- build computer-use agents that interact with any application

## When to choose
- you need deterministic, structured UI element references for agent-driven desktop automation
- you want to avoid brittle screenshot/pixel-matching approaches
- you're on macOS and want a fast single-binary CLI with MCP support

## When to avoid
- you need cross-platform desktop automation beyond macOS
- you're automating web apps, where browser automation tools fit better
- you need pixel-level visual verification of the UI

## Facets
- artifact type: cli-tool
- maturity: active
- function: cli, workflow-automation, mcp, developer-tools
- domain: developer-tools, accessibility, operating-systems
- platform: cli, rust
- tags: desktop-automation, accessibility-tree, ai-agents, mcp-server, gui-automation, structured-output, automation, macos

## Member repositories
- lahfir/agent-desktop (main) score 82

## Provenance
- Observed fields: from GitHub, fetched 2026-09-03T02:15:14.265387+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:07:47.413711+00:00, confidence not recorded.
  - readme: https://github.com/lahfir/agent-desktop (fetched 2026-09-03T02:15:14.265387+00:00, sha dd9981501b7c)
- Data as of 2026-08-30T08:39:29.467469+00:00.