What is LayaPilot for?

Enterprise admin-console testing: it observes live browser pages and generates replayable Excel test cases, or reads existing Excel cases and replays them against the app, covering table, filter, form, and CRUD-heavy pages.

How does the decision model fit in?

Playwright drives the browser while local Laya or a compatible API understands field purpose, picks form data strategies and dropdown categories, and matches replay targets. Assertions still come from page state, not model judgment.

What does it need to run?

Node.js 20 or newer, Python 3.10 or newer (3.11 or 3.12 recommended), and Playwright Chromium. Local model mode wants 4 or more cores, 16 GB RAM, and about 5 GB free disk, with no discrete GPU required.

How do I try the public demo?

Run npm run demo in one terminal, then run the generate command against http://127.0.0.1:8765/customers and execute the generated workbook with the execute command. Generate mode really clicks create, update, and delete, so use a dedicated test environment.

It is not a guarantee for arbitrary sites or spreadsheets: canvas, closed Shadow DOM, and custom controls need new adapters, and approvals, multi-account permissions, and file import still require dedicated work. Only Playwright and Laya are implemented today.

Where do results and logs go?

runs/ /执行结果.xlsx, results.json, and decisions.ndjson, which records form-field, form-option, test-discovery, and replay-target decisions. checkpoint.json and progress.ndjson are kept, but progress is for review and does not auto-resume.

How are credentials handled?

Passwords and API keys go only through environment variables, .env.local, or a hidden prompt, never into JSON config; config.local.json is gitignored. Generated workbooks and reports can contain target URLs or test data, so review them before sharing.

Jev directory

CommunityJev Tools & CLIs

LayaPilot

An enterprise admin-console test tool that generates replayable Excel test cases from live browser pages, or replays existing Excel cases, driving Playwright with local Laya or a compatible decision API to match fields, options, and targets.

See highlights Open github.com

Stars: 58
Language: JavaScript
License: MIT
Reading time: 2 min
Published by: Community
Author: anthonyli
Last verified: Oct 2026

anthonyli/laya-pilot

User repository on GitHub

View on GitHub

Generate replayable Excel test cases from live browser pages, or run existing Excel cases, using Playwright and local Laya or a compatible decision API.

Forks: 17
Last pushed: Updated Sep 2026

Repo stats from the GitHub API, cached Oct 2026.

Tagscommunity catalog-import madewithjev testing playwright excel

Works forJev for Code Jev for Verification

Used inJev for Software & DevTools

Languages: JavaScript Python
Platforms: Playwright CLI
Capabilities: Open sourceRuns locallySelf-hosted

Highlights

Generate mode writes a 17-column Excel workbook only for flows that really executed and passed assertions; hidden replay steps must match the visible cases.
Replay re-locates DOM elements and matches clicks, forms, dates, and dropdowns through the decision model; saved text, dates, and options replay exactly.
Three entry modes: --mode generate, --mode execute, and --excel for existing natural-language workbooks, which default to rows whose source result is pass.
Recommended local model is convaiinnovations/laya-multilingual for Chinese and English pages; the author validated on an Apple Silicon M4 with 16 GB.
Current scope is the Playwright driver plus the Laya decision protocol only; browser-use and the Jev API are roadmap items, not config switches.

Quickstart

bash

npm run demo
./run.sh --mode generate --url http://127.0.0.1:8765/customers \
  --case-file generated-cases/customers.xlsx
./run.sh --mode execute --case-file generated-cases/customers.xlsx \
  --url http://127.0.0.1:8765/customers

Watch out

README reviewed 2026-10-15. The repo itself lists its bounds: table, filter, and form CRUD pages, not arbitrary sites or Excel files, with approvals, multi-account permissions, and file import still needing adapters. The M4 timing and 16 GB advice are the author's setup rather than official minimums.

Questions about this entry

Short answers drawn from the linked project or page.

From the Jev guides

Original write-ups on what this entry is for.

Jev in Software Testing

Oct 11, 2026

A test loop is a long series of small judgments. What the Playwright, mobile QA, and CI experiments report: two calls per probe, 131 ms medians, and near coin-flip failures to design around.

Jev Practices & Patterns#testing#qa#confidence

Read article

Jev at home: voice pipelines, Assist routing, and sentinels

Oct 16, 2026

How Home Assistant conversation agents, macOS voice pipelines, and an advisory sentinel put a fast typed decision in the loop — one call per utterance, thresholds in code, and a daily token budget.

Jev Tools & CLIs#home-assistant#voice#automation

Read article

Search visibility, audited by decisions

Oct 16, 2026

Jev-based SEO tools split the crawl from the call: deterministic crawlers count and parse, while typed judgments rank fixes, place internal links, score search results, and track AI-answer mentions — with costs and caveats attached.

Jev Tools & CLIs#seo#search-visibility#audits

Read article

From the Jev AI model community

Posts from builders shipping with Jev right now.

Follow @typesafeai

LayaPilot

Highlights

Quickstart

Questions about this entry

What is LayaPilot for?

How does the decision model fit in?

What does it need to run?

How do I try the public demo?

What does it not do?

Where do results and logs go?

How are credentials handled?

More Jev resources like this

From the Jev guides

From the Jev AI model community