Open-source developer tools and skills for AI coding agents
HelixLab is a growing collection of developer tools that work with any AI coding agent. Each tool is a self-contained set of bash scripts and markdown workflow guides — no dependencies beyond standard Unix tools and ffmpeg.
Important
ffmpeg must be installed before using any HelixLab tools. Without it, video analysis scripts will fail immediately.
The setup skill can detect your OS and handle installation automatically:
/helixlab:setup
Or run manually: bash scripts/setup.sh
| Platform | Install Command |
|---|---|
| macOS | Run bash scripts/setup.sh (downloads static build with all filters) |
| Debian / Ubuntu | sudo apt install ffmpeg bc |
| RHEL / Fedora | sudo dnf install ffmpeg bc |
| Arch Linux | sudo pacman -S ffmpeg bc |
| Windows | Requires WSL2: sudo apt install ffmpeg bc |
Note
HelixLab requires ffmpeg with drawtext filter support for timestamp overlays.
The standard brew install ffmpeg on macOS does not include this.
The setup script downloads a static build
with all required filters included.
Verify installation:
bash tests/test-scripts.sh --check-ffmpegAdd the marketplace and install:
/plugin marketplace add https://github.com/mrrobertkent/helixlab-plugin
/plugin install helixlab
Then invoke skills directly:
/helixlab:vision-replay <video-path> <analysis-prompt>
/helixlab:help
Any ffmpeg-supported format works: mp4, webm, mov, mkv, avi, flv, etc.
Tip
Use --scope project when installing to share the plugin configuration with teammates via Git:
/plugin install helixlab --scope project
Other AI Agents
[!NOTE] All of the agents below use file-based configuration (AGENTS.md, GEMINI.md,
.cursor/rules/). Placing these files in your project root makes them project-scoped automatically.
HelixLab ships .cursor/rules/ for automatic context injection. Clone the repo and Cursor will detect the rules when video files are in context.
Alternatively, copy AGENTS.md to your project root — Cursor reads AGENTS.md natively.
Codex reads AGENTS.md automatically. Clone or copy AGENTS.md to your project root.
Gemini reads GEMINI.md automatically. Clone or copy GEMINI.md to your project root.
[!TIP] Gemini supports hierarchical config — a global
~/.gemini/GEMINI.mdplus a project-levelGEMINI.md. You can also configure Gemini CLI to readAGENTS.md:// ~/.gemini/settings.json { "context": { "fileName": ["AGENTS.md", "GEMINI.md"] } }
Kiro reads AGENTS.md from the project root. Clone or copy AGENTS.md to your project.
All of these agents support the AGENTS.md standard. Clone or copy AGENTS.md to your project root.
Manual (Any Agent)
All scripts work standalone — no plugin system required:
git clone https://github.com/mrrobertkent/helixlab-plugin.git
bash helixlab-plugin/skills/vision-replay/scripts/video-info.sh /path/to/video.mp4Tip
The setup script detects your OS and AI agent, checks dependencies, and provides tailored next steps.
bash scripts/setup.sh| Flag | Behavior |
|---|---|
--check |
Report dependency status only (no install prompts) |
--yes |
Skip prompts and auto-install missing dependencies |
Or use the skill in Claude Code:
/helixlab:setup
/helixlab:setup --check
Extract frames from video files using ffmpeg and analyze them with AI vision capabilities.
Three analysis modes:
| Mode | Use Case | Extraction |
|---|---|---|
| Animation Analysis | Timing, easing, smoothness, dropped frames | 10-60 fps |
| Page Load Analysis | Progressive rendering, FCP, LCP, layout shifts | Lighthouse-style intervals |
| Workflow Review | User journeys, state transitions, UX | 2-3 fps or scene detection |
Script quick start
# Get video info
bash skills/vision-replay/scripts/video-info.sh recording.webm
# Normalize (downscale + timestamp overlay)
bash skills/vision-replay/scripts/normalize-video.sh recording.webm /tmp/normalized.mp4
# Remove static/unchanged frames (threshold: 1=animation, 3=page-load, 15=workflow)
bash skills/vision-replay/scripts/dedupe-video.sh /tmp/normalized.mp4 /tmp/deduped.mp4 5
# Generate contact sheet overview
bash skills/vision-replay/scripts/contact-sheet.sh /tmp/deduped.mp4 /tmp/sheet.png
# Extract frames at 10fps
bash skills/vision-replay/scripts/extract-frames.sh /tmp/deduped.mp4 /tmp/frames 10
# Clean up
bash skills/vision-replay/scripts/cleanup.sh /tmp/framesLaunch a headed Chrome browser with built-in recording and annotation tools. Navigate manually while recording, draw annotations directly on the page, and save WebM recordings for AI analysis.
Note
Record Browser requires Node.js 22+. The install script downloads a self-contained Chrome for Testing binary — no npm packages or global browser install needed.
Features:
| Feature | Description |
|---|---|
| Glassmorphism toolbar | Record, On Action, pause/stop, restart |
| 6 drawing tools | Select, Pen, Line, Rectangle, Circle, Text (via fabric.js) |
| Lines & arrows | Plain line, arrow at end, arrow at start, double arrow |
| 5 color presets | Red, yellow, blue, green, white — optimized for AI vision |
| Stroke & fill | 4 stroke widths (1/2/3/5px), semi-transparent fill toggle for shapes |
| Text formatting | 4 sizes (S/M/L/XL), background toggle (BG), border toggle (BD) |
| Keyboard shortcuts | Space, D, Esc, Ctrl+Z, C, 1-4, Delete, Shift (constrain), ⌘/Ctrl+Click (multi-select) |
| Post-recording dialog | Video playback, expand overlay, rename, save & close |
| Welcome page | Launches when no URL provided — includes full tool reference |
How it works with Vision Replay:
Record Browser captures annotated recordings. Vision Replay analyzes them. The two skills chain together:
- Record:
/helixlab:record-browser→ navigate, annotate, save & close - Analyze:
/helixlab:vision-replay <saved-recording.webm>→ the agent sees your annotations as bright colored shapes in the extracted frames and focuses analysis on those areas
Quick start
# Install browser dependencies (one-time — downloads Chrome for Testing)
bash skills/record-browser/scripts/install-browser.sh
# Launch recorder (opens welcome page — navigate via address bar)
bash skills/record-browser/scripts/launch-recorder.sh
# Or launch directly at a URL
bash skills/record-browser/scripts/launch-recorder.sh "https://example.com"
# Use the in-browser toolbar to record, draw annotations, and save
# HELIX_SAVED=<path> is printed to stdout when the user saves a recordingPlatform support
| Platform | Status | Notes |
|---|---|---|
| macOS | Supported | Works out of the box |
| Linux (X11/Wayland) | Supported | Requires a display server |
| Windows (WSL2 + WSLg) | Supported | Windows 11 with WSLg provides GUI support automatically |
| Windows (WSL2, older) | Requires setup | Install an X server (VcXsrv, X410) and set DISPLAY |
Record Browser launches a headed (visible) Chrome window — it cannot run headless since the user needs to interact with the browser to navigate and annotate.
Troubleshooting
| Issue | Solution |
|---|---|
Chrome binary not found |
Run bash skills/record-browser/scripts/install-browser.sh |
Node.js 22+ is required |
Install Node.js 22+ from nodejs.org |
| Chrome window doesn't appear (WSL2) | Install WSLg (Windows 11) or an X server and set DISPLAY=:0 |
| Blank recording | Ensure you clicked Record or On Action before interacting |
| Annotations not visible in recording | Annotations are captured automatically — ensure Draw mode was active when drawing |
Note
Known Issue: Chrome for Testing v146 may show a "quit unexpectedly" dialog when the browser is closed. This is cosmetic and does not affect recording or video saving.
We welcome contributions! See CONTRIBUTING.md for full guidelines on:
- Reporting bugs and suggesting features
- Submitting pull requests
- Script conventions and skill structure requirements
- Commit message format and versioning
MIT — Robert Kent Jr.