Agent Teams

crewAI 1.14.5a5 is a pre-release update that deprecates CrewAgentExecutor in favor of AgentExecutor, improves Daytona sandbox tools, and addresses multiple security vulnerabilities in dependencies. The release includes bug fixes for HITL workflows, logging enhancements, and comprehensive documentation updates including migration guides for state restoration and crew-to-flow transitions.

OpenClaw v2026.5.12-beta.2 releases critical fixes and enhancements across authentication, provider streams, memory management, and UI/UX. Key improvements include auth-profile-backed media tools availability, subagent session visualization, error handling in auto-reply, and provider stream reliability. The release also upgrades to pnpm 11, adds per-agent tool policies, expands Slack/Gemini/Fal provider support, and strengthens TypeScript/linting standards.

OpenClaw v2026.5.12-beta.3 is a maintenance release addressing authentication, provider compatibility, and UI improvements. Key fixes include auth-profile-backed media tools availability, WhatsApp/Baileys dependency resolution, memory-wiki permission scoping, and provider stream handling. The release introduces per-agent tool policies, improved subagent session visualization, better error messaging, and expanded configuration options for Slack, Google Gemini, and local model providers.

OpenClaw v2026.5.12-beta.1 is a maintenance and feature release addressing security, UI/UX, and provider integrations. Key improvements include memory-wiki admin scope requirements, subagent session hierarchy visualization, Gemini 3 model normalization, enhanced tool policies, and expanded Slack/iMessage channel support. The release also upgrades build infrastructure to pnpm 11 and refines plugin SDK public APIs.

LangChain version 1.3.0 introduces support for `version="v3"` in the `stream_events` and `astream_events` methods, enabling enhanced event streaming capabilities for LangChain agents. This update allows developers to access improved event streaming functionality with better control and visibility into agent execution. The release includes comprehensive documentation in the event streaming guide for implementation details.

Claude Code v2.1.139 introduces significant new features including an agent view for managing multiple sessions, a `/goal` command for multi-turn task completion, and enhanced MCP server support with environment variables. The release includes numerous UX improvements like transcript navigation shortcuts, plugin token cost estimation, and fixes for critical issues such as credential deadlocks, memory leaks in MCP servers, and terminal interaction bugs.

OpenClaw v2026.5.10-beta.4 introduces significant enhancements across CI/CD, runtime detection, provider integrations, and agent capabilities. Key improvements include Fly Machines container environment detection, expanded Fal provider image editing support, stricter build tooling (oxlint, Vitest, TypeScript), and enhanced Slack integration with link/media unfurling controls. The release also upgrades to pnpm 11, adds per-agent message tool overrides, improves Discord voice diagnostics, and deprecates legacy Plugin SDK subpaths while exposing active model metadata.

OpenClaw v2026.5.10-beta.5 introduces significant improvements across CI/CD, runtime detection, provider integrations, and agent capabilities. Key updates include Fly Machines container detection, enhanced Fal provider image editing, stricter build tooling (oxlint, Vitest, TypeScript), pnpm 11 upgrade, and expanded agent-to-agent communication limits. The release also adds Slack thread management improvements, Discord voice diagnostics, plugin SDK deprecations, and QA automation enhancements for Telegram and Mantis.

OpenClaw v2026.5.10-beta.3 introduces stricter build and TypeScript checks, upgrades to pnpm 11, and adds significant enhancements across Slack, Discord voice, models, and the Plugin SDK. Key improvements include provider-level local service startup, context mapping, Slack link/media preview controls, realtime voice diagnostics, and structured media extraction capabilities. Multiple bug fixes address session management, authentication latency, and cross-platform compatibility.

OpenClaw v2026.5.10-beta.1 introduces significant enhancements to QA automation, real-time voice diagnostics, and gateway security. Key additions include Telegram live PR evidence automation with Crabbox integration, Discord voice diagnostics, and private skill archive uploads. The release also includes numerous bug fixes across CLI, agent routing, model compatibility, and chat delivery systems, plus dependency updates for improved performance and security.

crewAI releases version 1.14.5a4, a pre-release update that updates LLM listings and resolves a critical dependency issue by relocating the `textual` package to `crewai-cli` while adding `certifi` as a dependency. This patch addresses packaging conflicts and ensures proper CLI functionality. The release includes documentation updates to reflect the changelog and version information.

OpenClaw v2026.5.7 is a maintenance release focused on stability, security, and reliability improvements across plugin publishing, CLI tools, authentication, and multi-channel integrations. Key fixes address cron job persistence, Discord/Telegram/WhatsApp message routing, OAuth recovery, permission enforcement, and context caching issues. The release includes enhancements to CLI output formatting, voice capture quality, and error handling to prevent partial publishes and stale session states.

This guide teaches how to build a grade-and-revise loop using Outcomes in Claude Managed Agents, where a writer agent drafts a cited research brief and a stateless grader independently verifies every URL, quote, and claim against a detailed rubric. The grader provides structured feedback that drives revisions until the brief passes, eliminating manual review cycles. Key techniques include writing specific, actionable rubrics that force concrete evidence, using span.outcome_evaluation_* events to track the loop, and understanding when Outcomes is the right tool for quality assurance.

This cookbook demonstrates building a multiagent coordinator system using Claude Managed Agents to automate sales-proposal generation. A coordinator agent orchestrates three specialist subagents—a web-search researcher, a case-study librarian, and a rules-based pricing modeler—each with scoped toolsets and specific responsibilities. The pattern shows how to structure heterogeneous teams, use per-role tool scoping, and leverage thread_created/thread_message_received events for agent coordination to assemble tailored sales proposals.

OpenAI Agents Python SDK v0.16.0 introduces significant updates including a default model change from gpt-4.1 to gpt-5.4-mini, new configuration options for disabling max_turns limits, SDK-side tool execution concurrency control, and server-prefixed MCP tool naming to prevent conflicts. The release also includes stability fixes for tool call output indexes and security improvements for symlink handling. These changes enhance performance and provide greater flexibility for agent configuration while maintaining backward compatibility through explicit model specification.

LangChain v1.3.0a2 is an alpha release featuring improvements to stream events protocol (v3), ordered schema resolution, and agent middleware enhancements. The release includes performance optimizations, bug fixes across multiple integrations (OpenAI, Anthropic, Fireworks), and new features like HITL middleware respond decision and dynamic tool registration. Key updates involve dependency bumps, security patches, and refinements to the agent creation and model initialization systems.

crewAI v1.14.5a3 is a maintenance release focusing on bug fixes and architectural improvements. Key changes include correcting the status endpoint path structure, upgrading gitpython for security compliance, and extracting the CLI into a standalone package. This refactoring improves modularity while maintaining backward compatibility.

OpenAI Agents Python v0.15.2 introduces context management model settings and fixes critical issues with conversation session replay, tool execution, and error handling. The release includes 11 bug fixes addressing assistant conversation item IDs, function tool tracing, MCP tool metadata isolation, and stream terminal rejection. Comprehensive test coverage improvements ensure reliability across realtime tool behaviors, handoff mechanisms, and tool identity helpers.

This cookbook demonstrates building a vulnerability-discovery agent using the Claude Agent SDK that automatically threat-models C source code, hunts memory-safety bugs using built-in file tools (Read, Grep, Glob), and generates structured security reports. The agent operates in a multi-turn session with a bootstrap threat-modeling phase, an interview phase for owner input, and automated vulnerability finding and triage loops. The approach reduces false positives compared to traditional static analyzers by using Claude's reasoning to identify high-confidence memory-safety issues in a read-only sandbox environment.

crewAI v1.14.5a2 is a bug fix release addressing critical issues in task output handling, token counting, async operations, and LLM configuration. Key improvements include fixing task output restoration in finally blocks, preserving outputs across async batch flushes, and preventing incorrect final answer returns from hooks and errors. The release also enhances the CrewAIRagAdapter with proper kwargs forwarding and improves BaseModel input handling.