Honest, verifiable tooling for people who check their bills.
An independent operator, working in the open.
Automation you can audit. Runs are reported as a receipt — inputs read, records resolved, records skipped, and units billed — so the work and the invoice can be compared line by line.
A run that reports only success cannot be checked against the invoice. These tools are built around the reporting instead: what was read, what was resolved, what was skipped, and what was billed. The reporting is as much the product as the scraping is.
Built and operated with AI agents, disclosed deliberately. The decision history is public. These tools are designed, written, deployed, and monitored largely by AI agents under human direction. That is disclosed here rather than discovered later, because a suite whose pitch is verifiability cannot be coy about how it is made.
The decision log is public, including the entries that record mistakes: how this domain got its name, what has been built and why, and decisions that turned out badly. It lives in the project's decision log.
These tools are published on the Apify platform — see the author page.