Super excited to finally share Dynamic Workflows in Claude Code!! We built this a couple months ago, and it has slowly become a daily driver for a bunch of people at Anthropic. A few tips for getting the most out of it 🧵 x.com/ClaudeDevs/status/2060…

maximum intelligence

Happy Fable Day!

Within a few hours of using Fable it was clear that this model is on a different level. It has since changed the way I work, but more importantly it has changed the way I view the world and my place in it. - A large variety of tasks are now 1 shottable - It can debug large distributed systems BETTER than I can. Not just faster. Better. - Hard software design is now largely solved Don’t believe me. Try it for yourself

Dynamic workflows are the easiest, fastest and highest signal way to do code base wide bug sweeps. Also works for performance and security issues if you've closed the verification loop

Seeing a number of benchmarks showing Opus is the best model for long-running work. Five tips for running Opus autonomously for hours/days: 1. Use auto mode for permissions, so Claude doesn’t ask for approval 2. Use dynamic workflows, to have Claude orchestrate hundreds/thousands of agents to get a task done 3. Use /goal or /loop, to nudge Claude to keep going until it’s done 4. Use Claude Code in the cloud, so you can close your laptop (easiest way is the desktop or mobile app) 5. Make sure Claude has a way to self-verify its work end to end: Claude in Chrome browser extension for web, iOS/Android sim MCP for mobile, a way to start the full web server or service for backend work

Workflows have changed the way I use Claude Code. Once it clicks, it's hard to go back

I just turned one of our most token hungry skills into a dynamic workflow - now it uses upto 5x fewer output tokens while spawning more subagents! This means significantly lower AI costs for this job while increasing the reliability of the workflow purely from changes in the harness level. If this was running this on API pricing, savings would be massive - while increasing the reliability! Anthropic's insight of replacing an orchestrator from a non-deterministic AI agent to deterministic code written by an agent is such a simple and brilliant insight, it's blowing my mind! 🤯 So if you have tried to create skills to let your AI agent work do complex work, especially non-technical, it's time to ask your Claude to turn it into a Dynamic Workflow and see if that improves the outcomes.

Highly recommend trying out Dynamic Workflows if you get a chance. I also can't imagine working without them anymore (They are expensive to run, but you can ask claude to make them smaller to get a feel for token usage)

Yet another way to get your Claude to tackle more challenging tasks! Give /goal a shot A gnarly refactor I did yesterday ran for 14h straight

Finally a good way to manage your fleet of your Claudes from 1 place. Check it out and give us feedback!

Opus 4.7 uses more thinking tokens, so we've increased rate limits for all subscribers to make up for it. Enjoy!

For the developers building with Claude, a direct line from the team. Follow for changelogs, API releases, community updates, and deep dives.

Introducing Claude Opus 4.7, our most capable Opus model yet. It handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back. You can hand off your hardest work with less supervision.