LayaPilot skill
Skill for browser testing of admin panels: runs cases from Excel and collects a report with failure evidence.
- 63 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Drives a browser, stores test-stand credentials and may change data with the `--allow-write` flag.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/layapilotDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Drives a browser, stores test-stand credentials and may change data with the `--allow-write` flag.
This is third-party code. Review the repository files before installing.
What it does
LayaPilot takes ready test cases from Excel and runs them in a browser following the original steps and expected results. The outcome is a report with screenshots and evidence for failed checks. The main target is enterprise admin panels with tables, filters, forms and create, update and delete operations. The skill also works in reverse: it explores a page and derives new verified cases, read-only by default. The repository root is a complete skill with `SKILL.md` as the entry. The next-step decisions come from a model over a Chat Completions compatible API or a local model the installer downloads.
Who it is for. Testers and developers of enterprise admin panels who already have cases in Excel and do not want to rewrite them as code.
Good fit when
- A regression of dozens of Excel cases runs before release without manual clicking.
- A new module needs verified cases captured from a live page.
- A failed step is analyzed from `results.json` with its cause.
- Sign-in happens once and the state is kept for later runs.
Not a fit when
- The target is a public site or mobile app, while the main focus is admin panels.
- There is no API key and a local model cannot be installed.
- Tests must live in the repository as code rather than as a spreadsheet.
Example request
Use $laya-pilot on /absolute/path/cases.xlsx against the test stand. Check the file parsing first, then run headless and show the report with evidence.Limitations
Node.js 20 and Python 3.10 are required. The installer `node scripts/setup.mjs` sets up Playwright, Chromium and dependencies. A model makes the decisions: the API must be HTTPS Chat Completions compatible and return decision JSON of the required shape, since path compatibility alone is not enough. A local model needs a separate download, disk space and memory. Sign-in data sits in `.env.local`, which must not be committed. Steps that change data run only with the `--allow-write` flag.
How to disable. Delete the skill folder. The `.env.local` file, `.auth/` and the model in `models/` sit inside it and go with it.
FAQ
Can it run without an agent?
Yes, there are direct commands, for example `node scripts/run.mjs --excel cases.xlsx --url <address> --headless`. A workbook check without a browser uses `--inspect`.
What does generation mode do?
It explores a page and produces new cases verified in practice, usually read-only: `--mode generate --read-only`.
How to keep the sign-in?
Sign in manually or set credentials, then ask the agent to remember the state: the file goes to `.auth/` and is checked before the next run.
Is there a case template?
Yes, `assets/test-cases-template.xlsx` with three examples and a format description in `references/case-template.md`.
How to set up a local model?
`node scripts/setup.mjs --local-model`: by default `convaiinnovations/laya-multilingual` downloads into `models/laya-multilingual/`. Run with `--provider local`.
Discussion
No questions or reviews yet.
Related
SkillsA skill that lets an agent write and run Playwright scripts: page tests, forms, screenshots and link checks
A CLI and MCP for testing agent skills: repeatable YAML evals, report comparison, regression detection and skilled-improve
A library of 169 QA skills per language: test design, API, UI, regression, load, security and LLM testing.
A skill that guides Claude Code and Codex through writing and fixing pytest API tests with a real-run verification loop.
Thirteen testing skills from requirements to report: strategy, cases, API checks, Playwright and flaky-test triage.
A plugin of testing skills: from requirements and test cases to API, mobile and web runs backed by traceable evidence.
README in short
The skill reads Excel, drives a Playwright browser and writes a report. Docs cover architecture, retry policy and running a local model.