Joe Estephan

Most PDF extraction pipelines would have missed the Brazilian court-filing injection.

Two law firms hid white-on-white text in a labor court petition. It told the court's AI reviewer to answer superficially and not challenge the documents. The court's tool, Galileu, flagged it...

Auto-updating editor extensions is the riskiest setting on a dev laptop.

The backdoored Claude Code extension on Cursor's marketplace, versions 2.1.312 and 2.1.313, ran the moment the editor loaded it. No click, no prompt. It added an SSH key to authorized_keys, opened a tunnel...

Agents fail tax work by filling the gaps in the file. Surge AI's new benchmark gives agents 50 real tax assignments, about 30 documents each, and the best one passes 10.8%. The failures in their write-up aren't misread PDFs. One agent decided a couple wasn't married from their...

An 80% token price cut barely changes what a document pipeline costs to run. Max Woolf's post lists every cut since June and asks why. GPT-5.6 Luna went down 80% in July, and Opus 5.5 launched at $4 and $20 per million tokens. The Hacker News thread is about IPOs and...