agent-browser
Browser automation CLI for AI agents. Use for website interaction, form automation, screenshots, scraping, and web app verification. Prefer snapshot refs (@e1, @e2) for deterministic actions.
87
83%
Does it follow best practices?
Impact
90%
3.10xAverage score across 3 eval scenarios
Advisory
Suggest reviewing before use
Evaluation results
92%
67%Automate a Contact Form Submission
Deterministic ref loop and verification
Opens URL first
0%
100%
Snapshot -i before interaction
0%
100%
Uses @eN refs
0%
100%
Re-snapshot after action
0%
100%
networkidle wait used
0%
100%
No sleep commands
100%
100%
Verification evidence captured
0%
100%
Command chaining used
12%
75%
Install guard present
0%
0%
Script is executable bash
100%
100%
README present
100%
100%
82%
54%Concurrent Multi-Site Price Research Tool
Session isolation and cleanup
Named sessions used
0%
100%
Distinct sessions per site
0%
100%
All sessions closed
0%
100%
Closure on failure path
80%
100%
Snapshot -i per session
0%
0%
networkidle wait used
0%
100%
comparison.txt produced
100%
100%
Session design doc present
75%
100%
No shared session
70%
100%
session list used
0%
0%
97%
63%Authenticated Dashboard Scraper with Output Control
Auth state management and output control
state save used
0%
100%
state load used
0%
100%
Env vars for credentials
100%
100%
No hardcoded secrets
100%
100%
Focused snapshot flags
0%
100%
eval --stdin used
0%
100%
Deterministic ref loop in login
0%
100%
wait used after login
0%
100%
dashboard-data.txt produced
100%
50%
Automation notes present
100%
100%
Session closed
0%
100%
- Repository
- supercent-io/skills-template
- Commit
c033769
- 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.