Moxlade

We keep a few public data sources under continuous observation and serve them three ways: as Apify Actors, as an API, and as MCP servers for AI agents.

The difference from a scraper is time. A scrape answers "what is there now". A maintained corpus answers that, and also "what was there last month", "what changed", and "what disappeared" — questions that cannot be asked of a live page.

What that means in an Actor

• Results come from a table that is refreshed every day, so a run answers in seconds and cannot be blocked or rate-limited.
• Every row carries when it was read and how old it is.
• Filters run on the stored fields (score, hours, tenure, badges, rates), not only on what a search box accepts.
• A live fetch is available when you need the source's current answer; it lands in the corpus for the next person.
• You pay per delivered row. No start fee. A run that returns nothing costs nothing.

First source: Upwork freelancers — profiles, rates, earnings, Job Success Score, hours, contracts, availability. More sources follow the same shape.

Numbers on each listing are measured on our own acceptance runs, and every field in a README is one the Actor actually returns.

Questions and requests: support@moxlade.com