CtrlK
BlogDocsLog inGet started
Tessl Logo

computer-use-agents

Build AI agents that interact with computers like humans do - viewing screens, moving cursors, clicking buttons, and typing text. Covers Anthropic's Computer Use, OpenAI's Operator/CUA, and open-source alternatives.

52

Quality

58%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/computer-use-agents/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

51%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.

The body is a clean progressive-disclosure stub: concise and well-organized, correctly offloading detail to a single clearly-signaled reference. Its weakness is that it carries no actionable or sequenced guidance itself, so actionability and workflow clarity depend entirely on the referenced guide.

Suggestions

Add a brief inline quick-start or a 3-4 step high-level workflow with at least one validation checkpoint so the body is actionable without forcing a full read of the guide.

Include one small concrete example (e.g., a minimal computer-use tool call or a sandboxing safety check) in the body to raise actionability.

Surface the key safety/validation steps from the guide as a short inline checklist, since this is a critical-risk skill where validation must not be hidden behind a reference.

DimensionReasoningScore

Conciseness

The body is lean and well-sectioned with no padding or explanation of concepts Claude already knows; the only trim candidate is the opening paragraph repeating the description nearly verbatim, keeping it just below a 5.

4 / 5

Actionability

The body contains no concrete code, commands, or executable steps — it only routes the reader to the detailed guide ('Read the detailed guide before executing this skill'), matching the 'minimal concrete guidance; high-level hints' anchor; the actionable content lives in the reference file rather than the body.

2 / 5

Workflow Clarity

For a critical-risk, multi-step skill the body surfaces no sequenced workflow or validation checkpoints — it is a routing stub pointing at the guide; the rubric's cap-at-3 for missing validation applies, and the lack of even a rough in-body sequence pulls it to 2.

2 / 5

Progressive Disclosure

A clear, lean overview that signals a single one-level-deep reference (references/detailed-guide.md, confirmed to exist and be well-sectioned at ~2158 lines) with a markdown link and explicit read instructions, matching the 'clear overview with well-signaled one-level-deep references' anchor.

5 / 5

Total

13

/

20

Passed

Description

66%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.

The description clearly states what the skill does with concrete actions and named products, giving it strong specificity and distinctiveness. It is capped on completeness because it lacks any explicit 'Use when...' trigger guidance in the description itself.

Suggestions

Add an explicit 'Use when...' clause to the description naming the trigger phrases (e.g., 'Use when the user mentions computer use, desktop automation, screen control, or vision-based agents').

Include a few natural synonyms a user would actually say (browser agent, desktop automation, RPA) directly in the description rather than only in the body's 'When to Use' list.

DimensionReasoningScore

Specificity

Lists four concrete actions ('viewing screens, moving cursors, clicking buttons, and typing text') plus names the domain, matching the 'several specific actions; minor gaps' anchor; not a 5 because the framing 'interact with computers like humans do' is mildly abstract and the actions are generic to all computer-use rather than skill-specific.

4 / 5

Completeness

Has a clear 'what' (build agents that view/click/type) but no 'Use when...' or equivalent explicit trigger clause in the description, capping completeness at 3 per the rubric guideline.

3 / 5

Trigger Term Quality

Includes strong natural/product terms ('computer use', 'Computer Use', 'Operator/CUA') that a user would say, sitting above the 'missing common variations' anchor; not a 5 because natural synonyms a user might say (browser agent, desktop automation, RPA) appear only in the body, not in the description.

4 / 5

Distinctiveness Conflict Risk

Naming specific products (Anthropic Computer Use, OpenAI Operator/CUA) gives it a distinct niche with only minor overlap risk against generic automation/RPA skills; not a 5 because 'interact with computers like humans do' is broad enough to risk overlap with browser-automation or general agent skills.

4 / 5

Total

15

/

20

Passed

Validation

93%

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

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
sickn33/agentic-awesome-skills
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.