qa
Skill48 starsChanged 19 days agoMITView on GitHub
Launch an isolated Agentrium desktop instance for local manual QA alongside the user's production client.
In this repository10 other files it gives its agents.
- .codex/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- agentrium-conventionsSkill
- agentrium-conventionsSkill
- frontend-patternsSkill
- local-qaSkill
- rust-patternsSkill
- security-reviewSkill
- tdd-workflowSkill
- verification-loopSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Use this skill when asked to run, launch, or manually QA the desktop app.
- Very specific: line 14
Always set AGENTRIUM_INSTANCE_ID=.qa and pass all of these Tauri overrides together:
- Very specific: line 26
Use port 5174 by default, checking it first with Get-NetTCPConnection -LocalPort 5174 -State Listen -ErrorAction SilentlyContinue.
- Very specific: line 36
Wait for the Rust build and app launch.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Run Agentrium for local QA
Use this skill when asked to run, launch, or manually QA the desktop app. The desktop repository is the current directory; use its current branch and report any uncommitted changes before launch.
Keep the production client isolated
The user's production Agentrium may be running with live sessions. Never stop or close any existing Agentrium / claude-terminal.exe process, never use port 5173, and never run bare npm run tauri dev or npm run dev.
Always set AGENTRIUM_INSTANCE_ID=.qa and pass all of these Tauri overrides together:
- identifier
com.claudeterminal.desktop.qa - product name
Agentrium QA - first window title
Agentrium QA - Vite
devUrlon the selected QA port beforeDevCommandstarts Vite on that same port with--strictPort
The instance suffix isolates its database and keychain service. Git repositories, files, worktrees, and Claude Code's own session data remain real; direct destructive QA at a scratch repo.
Launch
Use port 5174 by default, checking it first with Get-NetTCPConnection -LocalPort 5174 -State Listen -ErrorAction SilentlyContinue. If occupied, choose the next free port above 5173; never kill a port owner. Install dependencies only if node_modules is absent.
On Windows PowerShell, write the Tauri JSON overrides to a temporary JSON file and pass that file to --config; quoting inline JSON through npm can strip its quotes. Set $env:AGENTRIUM_INSTANCE_ID = '.qa' in the same shell before running:
npm run tauri dev -- --features devtools --config .qa.tauri.conf.jsonUse the repository's vite.qa.config.ts for the QA server; it ignores src-tauri/target, whose locked executable otherwise causes Vite's watcher to exit with EBUSY. Set beforeDevCommand to npm run dev -- --config vite.qa.config.ts --port <QA_PORT> --strictPort. The Vite config is tracked and must remain in place while the server runs. Do not alter production ports or stop the watcher owner.
Wait for the Rust build and app launch. Verify using Get-Process claude-terminal | Select-Object Id,MainWindowTitle,StartTime; confirm a second process titled Agentrium QA exists and is newer than the pre-launch production process. If the isolated app cannot be confirmed, report the failure and stop without touching other processes.
Leave the QA instance running until the user says they are done. Tell them the branch, uncommitted changes, that the QA database and keychain are separate (they can retain data from previous QA runs), frontend edits hot-reload, Rust edits rebuild, and which external data remains real. Stop only the process/task started for QA when asked.
tauri dev --features devtools may rewrite src-tauri/Cargo.toml line endings or generated schema files. Report such side effects and do not commit them.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

