MCP
3 articles
How Claude Managed Agents Actually Work: Dreaming, Outcomes, Multiagent Orchestration, and Webhooks (2026)
Anthropic gave Claude Managed Agents four new mechanics at Code w/ Claude: Dreaming, Outcomes, Multiagent Orchestration, and Webhooks. The one that changes how you build is Outcomes — a separate grader that loops the agent until a rubric is met. Here is how each one works, and when to reach for it.
Harness Design for Long-Running AI Applications: Inside Anthropic's Generator-Evaluator Pattern (Claude Agent SDK, 2026)
On 24 March 2026 Anthropic Labs engineer Prithvi Rajasekaran published the most rigorous public account to date of how Anthropic designs harnesses for long-running AI applications — a GAN-inspired generator-evaluator pattern applied across two unusually different domains: frontend design (subjective, no binary verification) and full-stack coding (objective, machine-verifiable). The piece evolves the November 2025 Initializer + Coding Agent baseline into a three-agent planner + generator + evaluator architecture, with concrete cost-and-duration data ($200 / 6h on a retro game maker test, then $124 / 4h on a more ambitious DAW after the Opus 4.6 simplification pass). Inside the pattern, the two failure modes it fixes (context anxiety + self-evaluation bias), how it compares to LangGraph / AutoGen / OpenAI Assistants v2 / Devin, when it doesn't fit, and the canonical principle every team operating a harness should adopt: stress-test every component against the current model.
Building AI Agents in the Enterprise: Implementation Patterns for 2026
Anthropic's playbook is right about the enterprise shape. The missing layer is implementation: governed skills, MCP tools, memory, observability, worktree-safe orchestration, and agent fleets that survive contact with a 1,000-person company.