Works with the AI Agents you already use
Electron Runtime Verify
Unit tests and a green typecheck do not prove an Electron app works: the preload bridge, IPC handlers, menus, native dialogs and the packaged build only fail at runtime.
$2.99
Electron Runtime Verify
Example session with this skill installed
I added a "Move to Trash" menu item with a confirmation dialog to my Electron app. Check it actually works, and that the packaged macOS build still launches.
- Read your context and instructions
- Compiled the electron runtime verify
Built and launched the app unpackaged with a temp profile. Main process: Clean Up → Move to Trash is disabled with no selection and enabled after selecting a file. With dialogs stubbed: the confirmation shows the file name and size with Cancel as default; Cancel trashes nothing; Confirm calls shell.trashItem with the selected path and the item disappears from the list. Packaged arm64 build on a macOS runner: ad-hoc signature valid, fuses applied, smoke test passed (page loaded, preload bridge, 0 console errors) — screenshot attached. Not tested: real trashing (stubbed) and the Intel build (inspected only).
Connects securely to your tools. The creator never sees your data.
About this skill
Unit tests and a green typecheck do not prove an Electron app works: the preload bridge, IPC handlers, menus, native dialogs and the packaged build only fail at runtime. This skill drives the real app and checks it, with nothing but Node (22+ has a global WebSocket and fetch).
Two scripts do the talking (copy them into a scratch folder, never into the user's repo unless asked):
scripts/cdp.js— renderer: evaluate JS in the page, take screenshots, record console errors and network requests.scripts/main-inspect.js— main process: evaluate JS withrequire('electron')available (menus, windows, dialogs, IPC side effects).scripts/smoke-test.js— launches a packaged build and runs checks; exits non-zero on failure (made for CI).
Build the way the project does (read package.json scripts: Vite, Forge, electron-builder, webpack…). Then launch with a throwaway profile, so a single-instance lock or saved state from the user's own copy cannot interfere and nothing of theirs is touched:
Drives the app through debugging ports, so it cannot click inside native modal dialogs (it stubs them instead) and needs OS screenshots for native menus. Main-process inspection needs the unpackaged app when the inspect fuse is disabled. Does not sign or notarize builds.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 8 days ago
Requires
Permissions
Allowed Hosts
File Scopes
Talks only to the local app over localhost debugging ports. Destructive actions (trash, open file) are stubbed during tests; real file operations only in temp fixture folders.