Loading activity…
@claudedevs
Official updates for developers building with @ClaudeAI
AI
Anthropic says it is in discussions with multiple organizations about third-party evaluations of its safety practices.
AI
ClaudeDevs also shared a four-minute walkthrough to help newcomers get started with Claude Code Projects.
Technology
Google announced a persistent "Gemini agent" handling multi-day tasks across Workspace, Microsoft 365, Slack, and Jira. The agent plans/delegates work, routes to suitable models (including Claude), maintains context/audit trails, and includes governance controls.
Technology
SiliconANGLE reports Haiku 5.5 is a small model and says Sonnet 5.5 cache read prices were halved.
Technology
Anthropic says Haiku 5.5 costs around 75% less to run on average than Haiku 4.5.
AI
One observer says coding-agent strategies can feel like “YOLO” and finds it odd to learn about changes from random social media posts.
AI
ClaudeDevs describes output scanning designed to keep users informed about details they might miss.
AI
ClaudeDevs says the site features engineering deep dives, Claude Code and API guides, tips from the teams building Claude, and easter eggs.
AI
ClaudeDevs says its article offers guidance on evaluation design and skills Claude Code can use to improve applications.
AI
TestingCatalog reported on September 25, 2026, that the app was rolling out through Microsoft's Frontier program, bringing together Office Copilot, coding tools and a proactive assistant.
Technology
RuntimeWire, citing Anthropic, reports a 3.1x geometric-mean improvement across 13 performance measurements. That combines distinct tasks and products—not a threefold speedup for every interaction.
AI
Claude says it is also providing subscription users with a rate limit reset they can save and use whenever they choose.
Technology
ClaudeDevs says users can create a design review deck or UI mockup using files in their code repository, then edit it themselves or keep refining it in the conversation.
AI
ClaudeDevs says users can create test cases for a plugin or skill, score the runs, then rerun each case without the plugin to compare results.
AI
The team describes a Slack-triggered workflow: Claude pulls metrics, compares deployments and checks flags, then proposes a fix the team can approve and merge.
Technology
Anthropic just dropped Claude Fable 5.1, a powerhouse AI model that doubles down on tough coding challenges and long projects, making complex work feel more manageable.
AI
The AI assistant now performs desktop tasks while users work on other things.
AI
Updated model targets coding, reasoning, and long-running agentic tasks.
AI
Users react to the new concise output style in Claude Code and its per-project config behavior.
