Code with Claude: The 5 biggest updates explained

2026-05-11 · Source: Lenny's Newsletter · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, long

Summary

Anthropic's "Code with Claude" developer event introduced several key updates to its Claude code and Claude API products. These include "Routines," enabling scheduled or event-triggered actions via cron, HTTP, or GitHub webhooks, allowing automated tasks like weekly newsletter drafting based on a changelog. "Outcomes" in Claude manage agents allows users to define success rubrics for agents, which then iterate up to 20 times to meet specified goals. A new "Multi-Agent Framework" supports defining hierarchical teams of up to 25 agents, each with distinct toolsets, working collaboratively on a shared file system. "Dreams," currently in research preview, offers a primitive for agents to review past sessions and consolidate important memories. Additionally, Anthropic doubled Claude code's five-hour limits across Pro, Max, Team, and seat-based enterprise platforms, removed peak hour restrictions for Pro and Max plans, and increased rate limits for Opus models in the API.

Key takeaway

For AI Architects and VP of Engineering evaluating agentic platforms, Anthropic's latest updates position Claude as a strong contender for building sophisticated, autonomous workflows. The introduction of Routines, Outcomes, and a Multi-Agent Framework directly addresses common challenges in agent orchestration and reliability. You should explore these features to automate development tasks, enhance agent performance through defined rubrics, and scale complex projects with collaborative agent teams, leveraging the increased usage limits.

Key insights

Anthropic's new features enhance agent autonomy, collaboration, and memory, alongside increased usage limits.

Principles

• Define success with rubrics for agent self-correction.
• Orchestrate multi-agent teams for complex tasks.
• Consolidate agent memories for improved future performance.

Method

Define agent outcomes using a markdown rubric, allowing up to 20 iterations for self-correction. Orchestrate multi-agent teams via API, assigning roles and tools to up to 25 agents working on a shared file system.

In practice

• Automate weekly reports using Routines with cron triggers.
• Use Outcomes to refine PRDs against a defined rubric.
• Implement multi-agent teams for complex code reviews.

Topics

• Claude Code Routines
• Multi-Agent Orchestration
• Agent Goal Definition
• Agent Memory Management
• API Usage Limits

Best for: CTO, AI Architect, VP of Engineering/Data, AI Engineer, Machine Learning Engineer, AI Product Manager

Related on AIssential

• Live blog: Code w/ Claude 2026 — Anthropic is enhancing developer productivity and model autonomy through advanced agentic capabilities and platform integrations.
• How to Use Claude Managed Agents? — Anthropic's Claude Managed Agents simplifies AI agent deployment by providing a hosted, stateful, and secure execution environment.
• Claude Code Agent View Explained — Anthropic's Claude Code now features an intuitive interface for parallel multi-agent management.
• Claude Agent SDK [Full Workshop] — Thariq Shihipar, Anthropic — Anthropic's Claude Agent SDK simplifies building autonomous AI agents by packaging core components and emphasizing bash and file system use.
• 🤖 AI Agents Weekly: Claude Opus 4.8, Claude Code Dynamic Workflows, Chrome DevTools for Agents 1.0, DeepSWE, Agent Harness Scaling Laws, and More — The latest AI agent advancements focus on self-organizing systems and enhanced model judgment for more reliable, long-horizon autonomous operations.
• Claude Code, A Technical Guide for the Business Executive — Claude Code is transitioning from a developer tool to a secure, full-stack agentic OS, leveraging whole-project context.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Lenny's Newsletter.