Skip to content
Skills Library
Programming · Testing

oats-debugging

/oats-debugging

Use when reproducing or diagnosing runtime behavior in the running oats desktop app — driving and inspecting its windows via the oats-desktop MCP server (execute JS, read DOM, invoke backend commands). Pairs with systematic-debugging.

Repository
oats ↗
Licence
MIT
Last repository change
22.10.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 22.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use when reproducing or diagnosing runtime behavior in the running oats desktop app — driving and inspecting its windows via the oats-desktop MCP server (execute JS, read DOM, invoke backend commands). Pairs with systematic-debugging.

Original author description

Use when reproducing or diagnosing runtime behavior in the running oats desktop app — driving and inspecting its windows via the oats-desktop MCP server (execute JS, read DOM, invoke backend commands). Pairs with systematic-debugging.

How it works

  1. Enable the server
  2. Windows (what you'll see)
  3. executejs gotchas (these will bite you)
  4. Typical loop
Original SKILL.md headings · 49 lines, 343 words
SKILL.md preview
---
name: oats-debugging
description: Use when reproducing or diagnosing runtime behavior in the running oats desktop app — driving and inspecting its windows via the oats-desktop MCP server (execute JS, read DOM, invoke backend commands). Pairs with systematic-debugging.
---

# Debugging the running oats app via MCP

oats ships a Tauri MCP server (`tauri-plugin-mcp`) that lets you drive the live app —
list windows, screenshot, execute JS in a webview, and call backend commands. Use it to
verify a fix in the real app, not just in unit tests. Pair this with superpowers'
`systematic-debugging` for the methodology.

## Enable the server

It is registered in `.mcp.json` as **`oats-desktop`**. In Claude Code, it may be
disabled by default in `.claude/settings.local.json` (`disabledMcpjsonServers`);
remove it from that list or approve the server when prompted. The app must be launched
with the MCP feature:

```
npm run tauri:dev:debug      # = tauri dev --features mcp
```

Run it in the **background with the sandbox disabled** (cargo must compile and the GUI /
network must spawn). It's a never-exiting dev server — don't wait on it to "finish".

## Windows (what you'll see)

Routes live in `src/main.ts`. See `oats-architecture` for the full map. Quick version:
- `settings` — the settings window (pre-created hidden).
- `library` — titled **"Meetings"**; the main user-facing window.
- `main` — **headless/blank** BootstrapView; a blank screenshot of it is expected.

## execute_js gotchas (these will bite you)

- **Hidden WKWebViews are JS-suspended on macOS** — `execute_js` times out until the
  window is **shown and focused**. Show the window first.
- **Backend bridge**: `window.__TAURI__` is **not** exposed (the app uses ESM
  `@tauri-apps/api`). Call commands via
  `window.__TAURI_INTERNALS__.invoke('cmd_name', args)`.
- **No top-level `await`** — make the last expression a promise; the tool resolves it.
- A blank window with no JS bridge usually means the **Vue bundle failed to load** —
  check the vite dev log for an unresolved import (the router statically imports every
  view, so one bad import breaks the whole graph).

## Typical loop

1. Launch `tauri:dev:debug` in the background.
2. List windows; `show` the target window before any `execute_js`.
3. Inspect via JS / screenshot; reproduce the issue.
4. Drive the backend with `__TAURI_INTERNALS__.invoke(...)` to isolate frontend vs Rust.
5. Once root cause is found, fix per `oats-vue` / `oats-tauri` and re-verify here.

Versions

The database has no other skills named oats-debugging.

Database snapshot: 13.11.2026 · package version 52279bdaac95 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 90oats-architecture
    Use to orient in the oats codebase or plan cross-cutting features — the cloud-vs-offline backend split, the speech-to-text pipeline, window topology, and where specs live.
  • 100oats-tauri
    Use when changing oats backend — Rust commands in src-tauri, the invoke contract, capabilities/permissions, Tauri plugins, multi-window, or running the cargo test suite. Includes the macOS build/test workarounds.
  • 90Workflow Runtime Debug
    Debug workflow runtime behavior and test frontend features using the nc MCP bridge. Use when debugging workflow execution, testing UI components, inspecting DOM/state, writing frontend tests, or when user mentions nc_eval, nc_run_test, nc bridge, __workflowTest, test harness, or runtime debugging.
  • 100exploring-windows
    Explores a Windows desktop application (Notepad, Excel, an ERP client, a WinForms/WPF/Win32/UWP line-of-business app, a Save As dialog) through the robomotion-windows-mcp server - reads its UI tree like a DOM, searches it, takes screenshots, performs the steps - and returns a…
  • 88testing-windows-desktop
    End-to-end test the pi-gui Electron desktop app on Windows. Use when verifying Windows compatibility (dev launcher, PATH, folder picker, integrated terminal / node-pty) or any apps/desktop change on Windows.
  • 100next-dev-loop
    Verify Next.js runtime behavior after editing app code. Use this skill to confirm a change actually works in a running app — not just that it compiles or type-checks. Combines /_next/mcp (Next.js's view) with agent-browser (the browser's view). Requires a running `next dev`.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.