Anysite CLI

Command-line toolkit for web data extraction, batch processing, dataset pipelines, and analytics

What is Anysite CLI?

Anysite CLI is a command-line toolkit for web data extraction, batch processing, dataset pipelines, and analytics. It calls the same endpoints as the REST API, plus dataset pipelines, database loading, and LLM-powered analysis on top.

GitHub Repository

Source code, issues, and contributions — MIT licensed

Single API Calls

Make instant API requests with flexible output formats and field filtering

Batch Processing

Process thousands of inputs in parallel with rate limiting and error handling

Dataset Pipelines

Build declarative YAML workflows with chained dependencies and scheduling

Database Integration

Load data into SQLite, PostgreSQL, or ClickHouse with auto-schema

Database Discovery

Auto-discover database structure and browse saved catalogs

LLM Analysis

Classify, summarize, enrich, and deduplicate data using AI models

SQL Querying

Query collected datasets with DuckDB SQL and generate analytics

Agent Protocol

Structured JSON output, exit codes, and discovery for AI agents

Quick Start

Install the CLI

```bash
curl -fsSL https://raw.githubusercontent.com/anysiteio/anysite-cli/main/install.sh | bash
```
Or via pip: `pip install anysite-cli` — see [Installation](/docs/cli/installation) for all methods.

Configure your API key

```bash
anysite config set api_key YOUR_API_KEY
```

> ACTION FOR THE HUMAN: copy your API key at https://app.anysite.io/billing — an agent cannot obtain it.

Or log in via browser instead: `anysite auth login`. See [Authentication](/docs/authentication)
for where your key lives.

Update the schema cache

```bash
anysite schema update
```

Make your first request

```bash
anysite api /api/linkedin/user user=satyanadella
```

Key Capabilities

Feature	Description
Universal api command	Call any REST API endpoint by path with key=value parameters
Flexible Output	JSON, JSONL, CSV, and table formats with field selection and exclusion
Batch Processing	Parallel execution with rate limiting and configurable error handling
Dataset Pipelines	YAML-based multi-source workflows with dependency chains
Database Loading	Auto-schema creation, upsert, and diff-sync for SQLite, PostgreSQL, and ClickHouse
LLM Analysis	Classify, summarize, enrich, generate, match, and deduplicate records
Incremental Collection	Resume from where you left off with cursor tracking
Database Discovery	Auto-discover database structure, LLM-enriched catalogs
Agent Protocol	Structured errors, exit codes, and a discovery payload for scripts and agents

How It Fits in the Anysite Ecosystem

REST API

Direct HTTP API for programmatic integration into your applications

MCP Server

AI agent integration via Model Context Protocol for Claude, Cursor, and other tools

n8n Nodes

Visual workflow automation with drag-and-drop nodes

Want AI to build and run pipelines for you? See Claude Code Skill in the installation guide.