Agent Browser PRD Playbooks
Generates structured agent-browser CLI command sequences for verifying PRD user stories through automated UI walkthroughs.
The Agent Browser PRD Playbooks skill bridges the gap between product requirements and technical verification by generating ready-to-run automation scripts for the agent-browser tool. By parsing a prd.json file, it creates isolated, session-based command blocks tailored to specific user stories such as data enrichment, brand extraction, and system monitoring. This skill is essential for developers and QA engineers who need to quickly generate standardized evidence of functionality—including screenshots, URL assertions, and text verification—while ensuring a refs-first approach for stable browser interactions.
Key Features
01Ref-first command architecture for resilient and stable UI interactions
02Automatic playbook generation from structured PRD JSON files
03Automated evidence capture including full-page screenshots and metadata
040 GitHub stars
05Standardized templates for common flows like lead enrichment and monitoring
06Session-based isolation to prevent cookie and state leakage between tests
Use Cases
01Generating manual or automated UI walkthroughs to verify PRD compliance
02Bootstrapping browser automation scripts for new features in staging environments
03Capturing visual evidence and status reports for stakeholder reviews