peekaboo

Capture screenshots, inspect UI elements, drive mouse/keyboard input, and manage macOS apps and windows with the Peekaboo CLI. Use when automating macOS UI interactions, taking annotated screenshots, clicking UI elements by ID, typing text, controlling app windows, or navigating menus and Dock items.

Quality

91%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

High

Do not use without reviewing

The canonical home for this skill is peekaboo in Hung-Reo/hungreo-openclaw

Quality

Content

90%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 lean, highly actionable CLI reference: every example is executable, workflows are sequenced with a sensible pre-flight checkpoint, and depth is deferred to `--help`/`learn` rather than duplicated. The only structural gap is that the exhaustive command catalog lives inline in SKILL.md instead of a separate reference file, and post-action verification steps are absent.

Suggestions

Add a short error-recovery note to the main workflow (e.g., re-run `peekaboo see --annotate` to get fresh element IDs if a `click --on` fails because the snapshot is stale), giving an explicit validate-retry loop.

Move the full command catalog and common-parameters tables into a one-level-deep reference file (e.g. COMMANDS.md), keeping SKILL.md to the overview, quickstart, and flagship examples.

Include an explicit verification step after interactions (e.g., re-capture with `peekaboo image` or check `--json` output) so agents can confirm the intended UI state changed.

Dimension	Reasoning	Score
Conciseness	The body is dense and reference-style — one-line command catalog ("`click`: target by ID/query/coords with smart waits"), parameter tables, and runnable snippets — with no padding and no explanation of concepts Claude already knows. Not 4 because there is essentially nothing to trim; every section carries non-obvious CLI-specific information.	5 / 5
Actionability	Every example is a complete, copy-paste-ready command (see→click→type login flow, `--window-id` targeting, capture/analyze, menu/dock, mouse/keyboard), and the Quickstart is a runnable happy path. The examples cover the common cases the skill will be invoked for.	5 / 5
Workflow Clarity	Sequences are clear ("Quickstart (happy path)", "See -> click -> type (most reliable flow)") with a pre-flight checkpoint (`peekaboo permissions` first; "Use `peekaboo see --annotate` to identify targets before clicking"). Not 5 because there is no post-action verification or error-recovery guidance (e.g., what to do when a click target ID is stale); not 3 because the key checkpoint — snapshot before clicking — is explicitly present.	4 / 5
Progressive Disclosure	The actual bundle is a single file (no references/scripts/assets exist), and the skill wisely defers depth to the CLI itself ("`learn`: print the full agent guide + tool catalog", "Run `peekaboo <cmd> --help` for flags") with well-organized sections. Not 5 because at ~165 lines the full 30-command catalog and parameter tables are inline reference bulk that could be split into a one-level-deep reference file.	4 / 5
	Total	18 / 20 Passed

Description

92%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 description: concrete third-person capability statement paired with an explicit, natural-language "Use when" trigger list covering the main use cases. Trigger synonym coverage is good but not exhaustive, which is the only minor gap.

Dimension	Reasoning	Score
Specificity	"Capture screenshots, inspect UI elements, drive mouse/keyboard input, and manage macOS apps and windows" lists multiple concrete verb-led actions with comprehensive coverage of the CLI's capability space; not 4 because there are no meaningful coverage gaps, and it stays in third-person voice.	5 / 5
Completeness	Sentence one explicitly states what the skill does and sentence two provides an explicit "Use when..." clause with concrete trigger phrases, matching the top anchor's structure. Not 4 because the when-clause is already specific and concrete, not merely adequate.	5 / 5
Trigger Term Quality	Triggers like "automating macOS UI interactions, taking annotated screenshots, clicking UI elements by ID, typing text, controlling app windows, or navigating menus and Dock items" are natural phrases users would say, but a few common variations (e.g. "take a screenshot of a window", "press keys", "scroll") are missing. Not 3 because keyword coverage is broad and user-natural rather than partial; not 5 because synonym coverage is incomplete.	4 / 5
Distinctiveness Conflict Risk	The description carves a clear niche — macOS UI automation via a named CLI — and its triggers are platform- and tool-specific ("clicking UI elements by ID", "navigating menus and Dock items"), minimizing overlap with other skills. Not 4 because virtually no trigger term here would plausibly route to a different skill.	5 / 5
	Total	19 / 20 Passed

Validation

81%

Warnings & errors only

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

Validation — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	13 / 16 Passed

Repository: Bitterbot-AI/bitterbot-desktop
Path: skills/peekaboo/SKILL.md
Commit: 1c7dc1b

Reviewed: 20 hours ago

Table of Contents

Discovery Implementation Validation

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.