2 months ago
What should person been a regular merchandise has revealed immoderate of nan features Anthropic has been moving connected for Claude Code. As reported by Ars Technica, The Verge and others, aft nan institution released Claude Code's 2.1.88 update connected Tuesday, users recovered it contained a record that exposed nan app's root code. Before Anthropic took action to plug nan leak, nan codebase was uploaded to a nationalist GitHub repository, wherever it was subsequently copied much than 50,000 times. All told, nan full net (and Anthropic's competitors) sewage a chance to analyse much than 512,000 lines of codification and 2,000 TypeScript files.
In nan aftermath, immoderate group declare to person recovered grounds of upcoming features Anthropic is moving to develop. Over connected X, Alex Finn, nan laminitis of AI startup Creator Buddy, says he recovered a emblem for a characteristic called Proactive mode that will spot Claude Code activity moreover erstwhile nan personification hasn't prompted it to do something. Finn claims he besides recovered grounds of a crypto-based costs strategy that could perchance let AI agents to make autonomous payments. In a Reddit station spotted by The Verge, different personification recovered grounds that Anthropic mightiness person been moving connected a Tamagotchi-like virtual companion that "reacts to your coding" arsenic a benignant of April Fools joke.
"A Claude Code merchandise included immoderate soul root code. No delicate customer information aliases credentials were progressive aliases exposed," an Anthropic spokesperson told Bleepingcomputer. "This was a merchandise packaging rumor caused by quality error, not a information breach. We're rolling retired measures to forestall this from happening again."
As pinch immoderate different leak, it's worthy remembering plans tin and often do change. Just because a institution has written nan codification to support a characteristic doesn't mean it will yet vessel said feature.
