Posts tagged 'Anthropic'.
Anthropic added webhooks to Claude Managed Agents. Long-running AI agents now proactively notify your app on key state changes β idle waiting for approval, outcome evaluation complete, errors. With HMAC signing, replay protection, and SDK helpers in 7 languages, this is the practical guide for developers running AI agents in production.
In May 2026, Anthropic unveiled Claude Code 2.1 at the "Code with Claude" event. The /goal command lets Claude work autonomously for hours or days until a defined completion condition is met. Agent View gives developers a unified dashboard to manage dozens of parallel sessions. This is the moment AI coding tools cross from "assistant" to "autonomous system."
On May 11, 2026, Anthropic shipped Claude Code v2.1.139. The two headline features are Agent View and the /goal command. Agent View is a dashboard that surfaces all your sessions in one list. /goal lets you declare a completion condition and Claude keeps iterating until it meets it. AI coding tools are crossing from "does what I say" to "reaches what I want."
In May 2026, Claude Code's Hooks system received a significant upgrade. PostToolUse hooks can now intercept and replace output from any tool (not just MCP), hooks can read the current effort level via effort.level or $CLAUDE_EFFORT, and duration_ms tracks actual tool execution time. A new claude ultrareview subcommand also enables non-interactive code review from CI pipelines.
On May 6, 2026, Anthropic added three major capabilities to Claude Managed Agents: Dreaming (self-improving agents that review past sessions), Outcomes (rubric-based grading by a separate evaluator model), and Multi-Agent Orchestration (a lead agent that delegates tasks to specialist sub-agents). AI is evolving from assistant to self-improving teammate.