AI
LiteLLM 1.107.0 on PyPI and Docker carried a credential stealer
An attacker drove LiteLLM's own release pipeline with stolen GitHub tokens; pin 1.106.2, check for three dropped files and rotate keys.
Nina Park
Nina writes about AI, developer tools and the software shaping how engineers work. She cuts through the hype to what actually ships.
AI
Anthropic IPO draft reportedly shows $16.3B revenue, $80.7B loss
Most of the reported first-half loss is non-cash financing charges, and Anthropic's S-1 still isn't public.
AI
DeepSeek federal ban covers civilian agencies, not the public
DHS signed the order in July and made it public Monday; it reaches tools built on DeepSeek but never mentions open-weight models.
AI
Gemini 4 Argon vs Opus 5.5 vs GPT-6 Astra: benchmarks, price, limits
Argon costs half as much as Opus 5.5 through Nov. 30 and guesses far less in independent tests, but Pro users say its quota runs out fast.
AI
OpenCode says Console breach may have exposed users' AI provider keys
The breach hit OpenCode's hosted Console; its team says local CLI keys weren't involved, and affected users are told to revoke keys at each provider.
AI
How Claude traced the NIST curve seeds, and why it matters for HTTPS
Anthropic says Claude traced them to a 1990s C random-number generator seeded with the clock, leaving little room for a hidden backdoor.
AI
DeepSeek refreshes V4.1 Flash, claims coding and agent gains
The MIT-licensed update keeps the same API price, but no independent lab has tested it yet and running it locally takes datacenter hardware.