Sluicer

Sluicer is a local MCP server that extracts the structured data a web page declares and reports where each value came from, without a model or API key.

What you can do

• Extract the structured data a web page declares, such as embedded metadata and markup-level data.
• Return provenance for each extracted value, showing where on the page it came from.
• Run extraction locally without calling a model or requiring an API key.
• Fetch pages by default only from public addresses, with an opt-in switch for localhost and private networks.

When to use it

Useful when an assistant needs machine-readable facts from a web page and you want to know the source of each value rather than a model's paraphrase. Suited to structured-data extraction and verification tasks where no LLM or API key should be involved.

Requirements

Runs as a local process over stdio, installed from PyPI as sluicer and started with uvx, so a Python/uv environment is needed. No account, API key or authentication is declared. Network access is needed to fetch pages. The optional environment variable SLUICER_ALLOW_PRIVATE must be set to 1 to allow localhost and private-network fetches.