A researcher trained a cheap Gemma 2B judge to detect misalignment in AI agents, but testing against Anthropic's AuditBench showed the judge failed to reduce audit costs or reliably distinguish misali…
Anthropic shipped Claude Fable 5, a powerful Mythos-class model, on June 9, but the US Commerce Department placed it under export controls three days later, citing a jailbreak vulnerability. Anthropic…
AWS confirmed that a disruption affecting one Claude model on its Bedrock platform did not impact other Anthropic models, thanks to the platform's isolation architecture. The outage, attributed to hig…
Anthropic released Claude Fable 5, its most capable publicly available model, as part of a new Mythos-class tier designed for complex, multi-step reasoning and agentic workflows. The model features an…
Anthropic's 2026 Claude model lineup includes four tiers — Fable 5, Opus, Sonnet, and Haiku — each designed for specific task complexities and cost profiles. Teams that implement intelligent model rou…
Anthropic shipped Claude Fable 5, its first Mythos-class model, this week, posting a more than 10% benchmark improvement over Opus but blocking prompts related to cybersecurity, biology, chemistry, an…
Developer Joe Pohnelt created an email triage system using Claude Code subagents to automate Gmail management. The system uses six specialized AI agents for security analysis, relationship mapping, co…
Prism has released v1.4 Policy + Governance, a new layer that prevents AI cost overruns before they occur. The update introduces per-project policies that can deny specific models, force model selecti…
Ten headache specialists across the United States and Canada preferred expert-written summaries over those generated by three state-of-the-art large language models in a blinded evaluation of clinical…
A developer configured Claude Code to use Opus as an orchestrator and Sonnet as implementation and review subagents, keeping the main session focused and cost-efficient. The setup delegates heavy codi…
Developers are combining Claude Opus for planning and reasoning with Gemini 3.5 Flash for UI generation in a single multi-step coding workflow, routing each task to the model best suited for it. Claud…
A developer has released a free tool that tracks token spending in Claude Code sessions, alongside a paid kit containing 18 cost-reduction tactics. The free component includes a read-only script that …
Conductor released version 0.62.0 with file-backed settings in TOML format, an experimental browser preview powered by Agentation, and queue improvements with keyboard navigation. The update also intr…
A law firm reduced document review costs by 50% by configuring Claude Code and Cowork to use a sub-agent architecture. The setup assigns Opus 4.8 with maximum effort solely for content review while of…
A senior news editor ran Spec Kit's pipeline against a three-year-old personal CLI codebase and found that the tool's Constitution-generation step produced six mandates, four of which matched real arc…
A developer has identified nine strategies to reduce token consumption in Claude Code, emphasizing that excessive token use is a workflow problem rather than a pricing issue. The engineer's approach c…
Claude-code-replay reconstructs lost project state from Claude Code session logs by replaying file writes and optionally classifying Bash events through Claude Sonnet 4.6. The tool walks JSONL log fil…
Anthropic's Claude Code subscription tiers—Pro ($20), Max ($100), and Max ($200)—require fundamentally different workflows due to budget psychology, not just model access. On Pro, developers must rati…
GitHub reduced token usage in its agentic workflows by up to 62% after pruning unused Model Context Protocol tools, replacing MCP calls with GitHub CLI commands, and deploying daily audit and optimiza…
Anthropic released Claude Opus 4.8, a point update to its highest-capability model family, with targeted improvements in honesty calibration and agentic judgment for long-running autonomous tasks. The…