# LakeXpress — AI Agent Knowledge Base

Machine-readable data about LakeXpress, published for AI agents, LLMs, RAG systems, search
engines and other automated consumers.

**Base URL:** https://lakexpress.arpe.io/agents/
**Discovery file:** https://lakexpress.arpe.io/llms.txt
**Documentation for LLMs:** https://lakexpress-docs.arpe.io/llms.txt
**Product version covered:** LakeXpress 0.4.10
**Last updated:** 2026-10-22

## What LakeXpress is

A command-line tool that exports database tables to Parquet files, in parallel, on local disk
or cloud storage (S3, GCS, Azure, OneLake), and optionally publishes them as tables in
Snowflake, Databricks, Microsoft Fabric, BigQuery, Amazon Redshift, MotherDuck, AWS Glue or
DuckLake. Data is read by the FastBCP engine. Pipelines are stored as sync configurations in
the LakeXpress DB and run with `LakeXpress sync`, fully or incrementally.

## Files

| File | Contents |
| --- | --- |
| `manifest.json` | Index of every resource, plus context-selection rules, command rules and citation guidance |
| `product-overview.json` | Product identity, architecture, relation with FastBCP, differentiators, audience |
| `features.json` | Feature catalog with technical details for each capability |
| `benchmarks.json` | The customer results shown on the website, what is not published, performance levers |
| `supported-platforms.json` | Sources, storage backends, publishing targets, LakeXpress DB backends, operating systems |
| `use-cases.json` | Scenarios with problem, solution and complete commands |
| `pricing.json` | Plans from the pricing page, trial, licence resolution, FAQ |
| `technical-specs.json` | Commands, options, credentials file, export methods, exit codes, logging |

## Fetching

No authentication, no rate limit beyond standard web server limits.

```bash
curl https://lakexpress.arpe.io/agents/manifest.json
```

`manifest.json` lists every other file, so an agent can start there and follow
`resources[].file` to fetch only what a question needs. `aiGuidance.contextSelection` maps
common question types to the files worth loading.

## For assistants that build LakeXpress commands

These files summarise the product. To build or check real command lines, use:

- the **agent skill** `lakexpress`: https://github.com/arpe-io/arpeio-skills/tree/main/plugins/lakexpress
  (Claude Code: `/plugin install lakexpress --marketplace arpe-io/arpeio-skills`)
- the **MCP server** `arpeio-mcp`, hosted at `https://arpe-io-arpeio-mcp.hf.space/sse`, or
  `pip install arpeio-mcp` for local execution: https://github.com/arpe-io/arpeio-mcp
- the documentation: https://lakexpress-docs.arpe.io/latest/ (whole site in one file:
  https://lakexpress-docs.arpe.io/llms-full.txt)

## A note on performance figures

arpe.io does not publish a benchmark suite for LakeXpress. `benchmarks.json` carries only the
three customer results shown on the website, with their context, and says what is not
published. Please do not turn them into general throughput claims.

## Update policy

- Updated as needed with LakeXpress releases. See `CHANGELOG.md`.
- Semantic versioning on `manifest.version`.
- Additive changes only: existing fields are never removed.

## Attribution

Source: LakeXpress (https://lakexpress.arpe.io), Architecture & Performance
(https://www.arpe.io). Reuse is allowed with attribution; please cite the source URL and the
`lastUpdated` field of the file you used.

Questions: support@arpe.io