erhe-headless-verify
Automates E2E verification of editor behavior using headless Vulkan builds and programmatic MCP scripts.
Install
mkdir -p .claude/skills/erhe-headless-verify && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/16874" && unzip -o skill.zip -d .claude/skills/erhe-headless-verify && rm skill.zipInstalls to .claude/skills/erhe-headless-verify
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Verify editor behavior end-to-end without a display -- build the headless Vulkan editor, launch it, wait for its embedded MCP server, drive it with scripts/mcp_call.py (scene queries and mutations, geometry graph tools, undo/redo, capture_screenshot), read the screenshot PNG to SEE the result, then clean up. Use this whenever a change needs runtime verification and no interactive UI is required or no live display is available -- "does the mesh render", "does save/load round-trip", "does undo restore state", "what does the viewport look like now". This is the standard verification loop for editor changes on Windows; the windowed build cannot capture screenshots (returns "Frame capture not available") and aborts at startup when the display is off/asleep.Key capabilities
- →Build the headless Vulkan editor
- →Launch the editor and wait for its MCP server
- →Drive the editor with `scripts/mcp_call.py`
- →Read screenshot PNGs to verify visual results
- →Clean up editor processes and temporary files
- →Verify editor behavior end-to-end without a display
How it works
The skill builds and launches a headless Vulkan editor, drives it via an embedded MCP server using Python scripts, captures screenshots, and reads them to verify editor behavior without a display.
Inputs & outputs
When to use erhe-headless-verify
- →verify mesh rendering headlessly
- →test undo/redo functionality
- →verify save/load round-trip
About erhe-headless-verify
Builds a headless version of the editor to perform scene queries, mutations, and screenshot captures. Validates editor state through automated script execution.
Verify editor behavior end-to-end without a display -- build the headless Vulkan editor, launch it, wait for its embedded MCP server, drive it with scripts/mcp_call.py (scene queries and mutations, geometry graph tools, undo/redo, capture_screenshot), read the screenshot PNG to SEE the result, then
When not to use it
- →When an interactive UI is required
- →When a live display is available for verification
- →When the windowed build is sufficient for testing
Limitations
- →Requires a headless Vulkan build
- →Windowed build cannot capture screenshots
- →Requires manual cleanup of editor processes
How it compares
This skill enables end-to-end editor verification without a display by using a headless build, an embedded server, and programmatic screenshot capture, which is not possible with the windowed build.
Compared to similar skills
erhe-headless-verify side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| erhe-headless-verify (this skill) | 0 | 24d | Review | Advanced |
| amc-run-sample-calibration | 0 | 13d | Caution | Intermediate |
| webapp-testing | 353 | 3mo | Review | Intermediate |
| dev-browser | 53 | 4mo | Review | Intermediate |
Try saying
Example prompts that trigger this skill in your AI assistant.
You might also like
amc-run-sample-calibration
NVIDIA
Run end-to-end calibration on the shipped sample dataset (sdg_08_2_sample_data_010926.zip) against a running AMC microservice. Use when user says 'test sample dataset', 'run sample calibration', 'verify AMC install', or 'launch and test'.
webapp-testing
anthropics
Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.
dev-browser
SawyerHood
Browser automation with persistent page state. Use when users ask to navigate websites, fill forms, take screenshots, extract web data, test web apps, or automate browser workflows. Trigger phrases include "go to [url]", "click on", "fill out the form", "take a screenshot", "scrape", "automate", "test the website", "log into", or any browser interaction request.
playwright-browser-automation
lackeyjb
Complete browser automation with Playwright. Auto-detects dev servers, writes clean test scripts to /tmp. Test pages, fill forms, take screenshots, check responsive design, validate UX, test login flows, check links, automate any browser task. Use when user wants to test websites, automate browser interactions, validate web functionality, or perform any browser-based testing.
windows-ui-automation
martinholovsky
Expert in Windows UI Automation (UIA) and Win32 APIs for desktop automation. Specializes in accessible, secure automation of Windows applications including element discovery, input simulation, and process interaction. HIGH-RISK skill requiring strict security controls for system access.
qa-tester
svilupp
Browser automation QA testing skill. Systematically tests web applications for functionality, security, and usability issues. Reports findings by severity (CRITICAL/HIGH/MEDIUM/LOW) with immediate alerts for critical failures.