Claude Code Dynamic Workflows: Parallel Subagents & UltraCode Mode

Claude Code now supports dynamic workflows that let the agent orchestrate tens to hundreds of parallel subagents in a single session, checking work before anything reaches you. Available in research preview in the CLI, Desktop, and VS Code extension for Max, Team, and Enterprise plans, plus the API (Amazon Bedrock, Vertex AI, Microsoft Foundry).
How to use
- Ask directly: "Create a workflow"
- Enable UltraCode via the effort menu (sets effort to
xhigh), letting Claude auto-decide when to spawn a workflow.
Workflows consume substantially more tokens than a typical session — start scoped.
Capabilities shown in early access
- Codebase-wide bug hunts: parallel search across a repo, then independent verification on every finding.
- Large migrations: framework swaps, API deprecations, language ports spanning thousands of files end-to-end.
- Double-checked results: independent solution attempts and adversarial agents that try to break the result before delivering it.
Real-world example: Bun rewrite
Jarred Sumner used dynamic workflows to port Bun from Zig to Rust — ~750,000 lines of Rust in 11 days, with 99.8% of existing tests passing. Workflow steps:
- Map Rust lifetimes for every struct field.
- Write every
.rsfile via hundreds of parallel agents, each with two reviewers. - Fix loop driving build/test until clean.
- Overnight workflow to eliminate unnecessary data copies, opening a PR for each fix.
How it works
Claude plans dynamically from your prompt, breaks the task into subtasks, fans work across parallel subagents, checks results independently, and iterates until answers converge — producing a single coordinated answer.
📖 Read the full source: HN AI Agents
👀 See Also

agentmemory V4 achieves 96.2% on LongMemEval benchmark, outperforms commercial AI memory systems
agentmemory V4 scored 96.2% on LongMemEval, beating several funded AI memory companies including PwC Chronos (95.6%), Mastra (94.87%), and OMEGA (93.2%). The system was built solo in 16 days on a mid-range gaming PC with a $1,000 budget.

Local 35B MoE Model Drops Agent OS Code Failure Rate to 0%
A developer reports that switching a multi-agent system's runtime to Qwen 3.6 35B A3B (MoE, 3B active params) eliminated code failures, achieving 100% success rate through a 5-layer validation gate.

I design with Claude more than Figma now — Jane Street designer workflow
A Jane Street designer describes how Claude Code replaced Figma for prototyping, saving weeks of back-and-forth. Prototype features ship as real code, not mockups.

HostedShell: A Web-Based Deployment Solution for OpenClaw Agents
HostedShell is a hosted version of OpenClaw that eliminates local CLI setup, dependency management, and manual pairing by providing a web console with direct terminal access and filesystem updates.