Stats
Actions
Tags
How this skill is triggered — by the user, by Claude, or both
Slash command
/helixlab:helpThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
**Version:** 1.2.0
Version: 1.2.0 License: MIT Author: Robert Kent Jr. Homepage: https://github.com/mrrobertkent/helixlab-plugin
| Skill | Invoke | Description |
|---|---|---|
| Vision Replay | /helixlab:vision-replay <video-path> [instructions] | Extract frames from video files using ffmpeg and analyze them with AI vision. Supports animation timing analysis, page load performance review, and user workflow progression. |
| Record Browser | /helixlab:record-browser [url] | Launch a headed Chrome browser with recording controls and annotation tools. Draw annotations (lines, arrows, rectangles, circles, text) on the page — captured in the WebM recording for AI analysis with vision-replay. |
| Setup | /helixlab:setup [--check] | Check prerequisites and install dependencies (ffmpeg, ffprobe, bc). Detects your OS and AI coding agent for tailored setup guidance. |
bash scripts/setup.sh (downloads static build with drawtext support)sudo apt install ffmpeg bcsudo dnf install ffmpeg bcsudo apt install ffmpeg bcRun /helixlab:setup --check to verify all prerequisites are met.
/helixlab:vision-replay /path/to/recording.webm analyze the page load animation
The skill will:
npx claudepluginhub mrrobertkent/helixlab-plugin --plugin helixlabAnalyzes screen captures from the local screenpipe daemon, detects repeated research workflows, and drafts new skills or composition recipes for uncovered patterns. Use on-demand to turn work history into reusable skills.
Generates a polished demo video from a Playwright trace recording by parsing actions, writing voiceover scripts, and producing SRT subtitles.
Teaches skill invocation protocol, tool selection rules, agent delegation patterns, and enforcement mechanisms for all Claude Code sessions.