//Anysite CLI

Get a fresh B2B lead list in your database every week

Describe the companies and people you want in one YAML file. The open-source CLI collects them on a schedule.

Start for free→
anysite cli
  • Claude
  • Cursor
  • Manus
  • Grok
  • Gemini CLI
  • Codex
  • Antigravity
  • OpenClaw
  • OpenAgent
//Connect

From pip install to a lead table

Copy the commands as they are, or let your agent run them.

Install the CLI
//What you get

Your lead list, updated on a schedule

You write the YAML, or your agent does. The CLI runs it and keeps it fresh.

Fits in one YAML file

Set sources, filters, storage and schedule.

YAML${source.field}

Runs on a schedule

With --incremental, skip what was already collected, then fire a webhook.

CronIncrementalWebhook

Lands in your database

Query it all with SQL through DuckDB.

SQLitePostgreSQLClickHouse

Works with your agent

Let your agent write the YAML and run it.

--dry-runJSON output

Every command in the docs

//How it works

Install and add your key01Describe the list in YAML02Get a fresh lead table every week03

//Products

Other ways to use Anysite

Anysite MCP

Ask your agent for enriched B2B leads.

ClaudeCursorFrom $30/mo

Anysite API

Enrich leads and signups in your app.

RESTCLIFrom $49/mo

Anysite Datasets

Get company and people data in bulk.

70M+ companies850M+ profilesFree sample
//Q&A

Questions about the CLI

Explore in the chat with MCP. When the job repeats, run it as a scheduled YAML pipeline with the CLI. Both run on the same engine and the same 900+ sources.

An API plan, from $49 a month. The CLI uses your API key and its credits; MCP plans don’t cover it. Starter comes with a 7-day free trial and 1,000 credits. The CLI itself is free and open source under MIT.

SQLite for local work, PostgreSQL for shared tables, ClickHouse for analytics at volume, or plain Parquet files. Every destination stays queryable with SQL through DuckDB.

The anysite dataset schedule command generates the cron or systemd entry for you to install. Incremental runs skip what was already collected, and a webhook fires when the run completes.

No. Only the pipeline config enters the context. Collection, storage and SQL run locally, so records never pass through the model.

Put your lead list on a schedule