Explorbot CLI
Explorbot autonomously explores a web app in the browser, finds bugs and saves flows as runnable Playwright tests.
- 104 ★ on the repository
- Last commit
Broad access
Why this level
- The agent clicks, fills forms and may change or delete data in the tested app on its own.
- Page content and the credentials from `EXPLORBOT_AUTH` reach the AI provider requests and logs.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/explorbotDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- The agent clicks, fills forms and may change or delete data in the tested app on its own.
- Page content and the credentials from `EXPLORBOT_AUTH` reach the AI provider requests and logs.
This is third-party code. Review the repository files before installing.
What it does
Explorbot is a CLI with a crew of AI agents for exploratory testing. You supply a URL and a goal; the agents map the page, plan scenarios, drive the interface, verify outcomes and cluster findings by root cause. A run leaves behind Playwright or CodeceptJS tests, HTML and Markdown reports, step-by-step screencasts and experience reused on later runs. It works with any provider through the Vercel AI SDK and runs in CI from environment variables.
Who it is for. QA engineers and developers who want an exploratory layer on top of scripted tests, including in CI.
Good fit when
- An admin section is new and has no tests yet.
- You want overnight monkey-testing of an MVP for unexpected paths.
- Exploration should yield Playwright specs ready to commit.
- An agent calls Explorbot in CI with `--max-tests`.
Not a fit when
- The app is production and no staging environment exists: the agent changes data.
- Page content must not be sent to an external AI provider.
- You need deterministic regression checks, not exploration.
Example request
Explore the `/admin/users` section on the staging site, produce at most three tests and save the report.Limitations
You need Node.js 24 or Bun, an AI provider key (OpenRouter recommended) and a modern terminal. CRUD-heavy apps fit best. Login credentials are passed in `EXPLORBOT_AUTH` or the config: create dedicated test accounts. The repository license is listed as NOASSERTION, so check the terms before company use. The terminal must be interactive: a session is driven by commands such as `/explore`, while CI uses the separate `explore` mode.
How to disable. Remove the `explorbot` package from dependencies and the `~/.explorbot/sites/` directory with accumulated reports and experience. In CI drop the `EXPLORBOT_*` variables.
FAQ
How do I run it in CI?
Set `EXPLORBOT_URL`, `EXPLORBOT_AI_PROVIDER` and `EXPLORBOT_AUTH`, then run `npx explorbot explore /admin/users --max-tests 3`. No config file is needed.
Where do results go?
To `~/.explorbot/sites/<host>/`; nothing is written to your project. `EXPLORBOT_EPHEMERAL=1` stops anything being kept between runs.
Which AI provider should I choose?
The README recommends OpenRouter; Groq, Cerebras, OpenAI and Anthropic also work through the Vercel AI SDK. Recommended models are listed in `models.json`.
What tests does it produce?
Playwright or CodeceptJS specs for every passing flow, ready to commit and run in CI.
How much does a run cost?
The README promises cents per session: cheap models do the heavy work while smarter ones decide from short logs. The exact price depends on the provider.
Discussion
No questions or reviews yet.
Related
CLI toolsA skill that tests your agent's code in a real browser: reads your changes, builds a test plan and runs it through Playwright with your own cookies
Playwright CLI
playwright-cli
Microsoft's official Playwright CLI with an agent skill: drive a browser through short commands without heavy MCP schemas
A team of AI testers that uses your app, finds bugs, fixes them in a separate worktree and opens pull requests and issues.
A skill that lets an agent write and run Playwright scripts: page tests, forms, screenshots and link checks
Seven QA agents and ten skills for test automation: Playwright in TypeScript, Selenium in Java, API, accessibility and ISTQB.
Vercel CLI for testing coding agents on your framework: runs Claude Code, Codex and others and measures the success rate
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| Explorbot (this card) | no | 104 | ||
| Agentic Playwright | no | 283 | drives Playwright from your own agent session; Explorbot runs the exploration itself and keeps the tests | |
| Playwright MCP | yes | not required | 39,508 | gives an agent browser tools; Explorbot runs a ready crew of testing agents |
README in short
Install through npm, then `npx explorbot init` and `npx explorbot start`. A session follows research, plan, test, verify and keep. Output: tests, reports, videos and experience.