Understand this story
On October 22, 2026, Anthropic published an article offering a migration method for folding subagents into the main agent loop. Its applied AI team found in comparisons across multiple enterprise deployments that a single agent with skills often outperforms subagent architectures in quality, with lower per-task cost and latency. The article also notes that subagents remain suitable for scenarios where context bloat degrades performance, where independent tasks need to be processed in parallel, or where tool selection conflicts with instructions, and recommends first treating subagents as tool calls, then converting them one by one into skills and MCP tools, using end-to-end evaluations to compare performance before and after. The article mentions that Claude Code's system prompt for more advanced models was reduced by over 80%, and provides companion Claude API skills such as build-eval, prompt-audit, hillclimb, and cost-optimize. There are currently no further developments.
AI-generated from public reports
Latest development · Report
Anthropic shares when to fold subagents into the main agent loopDevelopments
1 confirmed developments
- Source:Claude Blog
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Oct 22
Claude BlogFirst-party
Anthropic shares when to fold subagents into the main agent loop
In comparisons across multiple enterprise deployments, Anthropic's Applied AI team found that a single agent with skills often outperforms subagent architectures in quality, with lower per-task cost and latency, and presents a migration method for folding subagents into the main loop. The article notes that subagents remain suitable for scenarios where context bloat degrades performance, independent tasks need to be processed in parallel, or tool selection conflicts with instructions, and recommends first treating subagents as tool calls, then converting them one by one into skills and MCP tools, using end-to-end evaluations to compare performance before and after. The article mentions that Claude Code's system prompt for more advanced models has been reduced by over 80%, accompanied by Claude API skills such as build-eval, prompt-audit, hillclimb, and cost-optimize.