Anthropic introduces “agent teams” in Claude Code as a research preview to launch multiple agents that work in parallel, and a Claude in PowerPoint plugin
On Thursday, Anthropic released the latest version of Opus — its most advanced model and a particularly important model for Claude Code.
Context & Ripple Effects
Anthropic had already positioned Opus as central to Claude Code, following its earlier claim that Opus 4.5 improved coding, agents and computer use. This release turns that model emphasis into product surfaces: coordinated parallel work in the coding tool and a presentation-software integration.
The subsequent Opus 4.7 update for advanced software engineering suggests the company is continuing to pair model upgrades with higher-effort technical workflows, rather than treating Claude Code as a standalone chat interface.
First-order effects
- Claude Code users can test a research-preview workflow that assigns segmented work to multiple coordinating agents in parallel, while Anthropic expands Claude into PowerPoint.
- The release makes Opus 4.6 more consequential for Claude Code, aligning with Anthropic's claim that the model can focus more deeply on difficult task components in its Opus 4.6 evaluation.
Second-order effects
- Coding-assistant vendors are pressured to compete on orchestration of multi-step work, not only on the quality of a single model response.
- Adding PowerPoint broadens Claude's relevance from software development into document-production workflows, increasing the value of a common assistant layer across different work surfaces.
Third-order effects
- If parallel-agent coordination proves reliable, AI coding products may evolve from single-assistant interfaces toward managed teams of specialized agents, with control and review becoming central product requirements.
- The combination of coding and presentation integrations points toward workplace AI competing through embedded workflow coverage; adoption will depend on whether users can govern output quality across those surfaces.
The trend: This is part of the shift from standalone AI assistants toward agentic work surfaces that coordinate tasks across the tools where knowledge work happens.