peekaboo
Capture and automate macOS UI with the Peekaboo CLI.
65
Quality
56%
Does it follow best practices?
Impact
77%
1.92xAverage score across 3 eval scenarios
Risky
Do not use without reviewing
Optimize this skill with Tessl
npx tessl skill review --optimize ./openclaw/skills/peekaboo/SKILL.mdEvaluation results
65%
35%Automate Employee Portal Login
See-click-type reliable flow
Annotate before click
0%
100%
See before click order
0%
100%
Element ID targeting
0%
0%
Clear before type
100%
0%
Return for submission
100%
100%
Screenshot path saved
50%
100%
App targeting
0%
100%
Permissions check
0%
0%
Without context: $0.1514 · 40s · 9 turns · 14 in / 2,230 out tokens
With context: $0.2370 · 51s · 14 turns · 17 in / 2,561 out tokens
80%
7%Window Inventory and Visual Report Generator
JSON scripting and capture parameters
JSON flag on list commands
53%
100%
Retina capture
100%
100%
AI analysis flag
100%
100%
Screenshot path output
100%
100%
Window-mode capture
100%
100%
polter invocation
0%
0%
App/window enumeration
100%
100%
Output format selection
0%
0%
Without context: $0.8526 · 3m 30s · 30 turns · 279 in / 8,861 out tokens
With context: $0.3736 · 1m 10s · 20 turns · 180 in / 4,423 out tokens
88%
71%Automated Kanban Board Reorganization
Gesture, keyboard, and motion parameters
Smooth scroll
0%
100%
Human movement profile
0%
0%
Drag duration
0%
100%
Hotkey comma syntax
0%
100%
Annotate before drag
0%
100%
Element ID for drag
0%
100%
Press with count
0%
100%
Scroll direction flag
100%
100%
Scroll amount
100%
100%
Without context: $0.3372 · 1m 22s · 19 turns · 24 in / 4,976 out tokens
With context: $0.4142 · 1m 30s · 20 turns · 26 in / 5,189 out tokens
- Repository
- trpc-group/trpc-agent-go
- Commit
8763418
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
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.