CtrlK
BlogDocsLog inGet started
Tessl Logo

host-computer-use-macos

Backend-specific macOS guidance for `computer_use_remote`. Load after `status` or `start_session` reports backend_family/backend_id `macos` or Accessibility-tree features. Covers AX structural targeting, macOS window actions, and screenshot verification.

69

Quality

84%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

88%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-crafted, executable skill body: concrete tool-call examples, explicit verification feedback loops, and clean sectioning. The only minor gaps are a few trimmable sentences and the lack of a short orienting overview at the top of the body.

Suggestions

Add a one- or two-line overview immediately after the heading summarizing the AX snapshot -> action -> verify loop, so the core workflow is visible before the detail.

Tighten the Permissions section by collapsing the repeated 'inspect the fresh screenshot' guidance (already stated in AX Targeting and Window Actions) into a single shared verification rule.

DimensionReasoningScore

Conciseness

The body is instruction-dense with minimal padding and assumes Claude's competence; explanatory lines like 'AX actions are attempts, not proof' earn their place, though a few sentences could be tightened further, keeping it just below a 5.

4 / 5

Actionability

Provides complete, copy-paste-ready JSON tool-call examples with real action/target/operation fields, enumerates supported operations (press, focus, set_value), concrete shortcuts (Command+H/M), and specific error codes (COMPUTER_USE_REARM_REQUIRED), covering the common cases fully.

5 / 5

Workflow Clarity

Sequences the flow (ax_snapshot -> choose element -> ax_action -> inspect screenshot) with explicit validation checkpoints and feedback loops for error recovery (re-snapshot on ambiguity or missing target, verify screenshot after state-changing actions), satisfying the top anchor including for destructive/state-changing operations.

5 / 5

Progressive Disclosure

Well-organized into three clear sections (AX Targeting, Window Actions, Permissions) with no nested references and no external files needed for this self-contained skill; a brief overview/summary at the top and slightly tighter section intros would lift it to a 5.

4 / 5

Total

18

/

20

Passed

Description

80%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, tightly-scoped description that clearly states both capabilities and a concrete load condition. Its only weakness is trigger-term naturalness: the triggers are technical/programmatic rather than user-spoken phrases, with no synonyms or variations.

Suggestions

Add a natural-language trigger phrase or synonym (e.g., 'macOS computer control', 'Accessibility automation') so the skill is discoverable from user-spoken terms, not only tool output.

Consider naming the file extensions or tool names a user might reference (e.g., 'A0 CLI', 'macOS GUI automation') to broaden keyword coverage.

DimensionReasoningScore

Specificity

Names the macOS computer-use domain and lists several concrete capability areas — 'AX structural targeting, macOS window actions, and screenshot verification' — with only minor coverage gaps, matching the 'lists several specific actions' anchor rather than the 1-2 actions of a 3.

4 / 5

Completeness

Explicitly answers both what ('Covers AX structural targeting, macOS window actions, and screenshot verification') and when ('Load after status or start_session reports backend_family/backend_id macos or Accessibility-tree features') with concrete trigger phrases, matching the top anchor.

5 / 5

Trigger Term Quality

Relevant domain keywords are present (macOS, Accessibility-tree, computer_use_remote, backend_family/backend_id) but they are technical/programmatic triggers rather than natural phrases a user would say, and no synonyms or common variations are included.

3 / 5

Distinctiveness Conflict Risk

A clear niche (macOS backend only) with a precise programmatic trigger that is hard to misfire, giving minimal conflict risk with other computer-use skills.

5 / 5

Total

17

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
agent0ai/agent-zero
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.