AI cuts coding time, but it's also filling open source's inbox with junk
Open-source maintainers estimate that a quarter of incoming pull requests are AI slop. On larger projects the share rises to 31%, adding to their review burden.
Related stories
Hermes Agent 0.22.3 rolls up about 620 merged changes, adding live MCP tool progress
Hermes Agent 0.22.3 is a patch release that bundles roughly 620 merged pull requests since the previous version into a stable build for the CLI, the Microsoft Store app, Docker and Hermes Cloud. Among the changes, MCP progress notifications now appear as live tool progress and the classic command-line layout has been refreshed; detailed notes are promised with version 0.23.0.
Granite Speech 5.0: transcribe, translate and attribute more accurately and faster
Granite Speech 5.0 has been presented on the Hugging Face blog. According to its title, the new version transcribes, translates and attributes speech more accurately and faster than before.
NVIDIA Agent Toolkit expands, grounding enterprise agents in company data
NVIDIA has added new tools to its open-source Agent Toolkit. Companies use it to customize and control agents, feed them internal data for their answers and actions, and monitor what they do.
PHP rushes out emergency patches for two critical bugs AI found first
PHP has shipped emergency updates fixing two critical flaws, found by AI, that let an unauthenticated attacker run code on almost any server using the language. The fixes, in versions 8.5.12, 8.4.27, 8.3.36 and 8.2.35, ended two frantic days in which websites around the internet were taken offline while waiting for a patch.
Anthropic set Claude on 25 open-source projects and it found 493 flaws
Anthropic pointed Claude at 25 widely used open-source projects, asking it to find code that attackers can reach when they should not be able to rather than isolated bugs. Claude flagged 493 such issues, one of which had been in the software for 21 years.