Tool Use

OpenAI Agents Python v0.17.5 release includes bug fixes for sandbox error retryability, tool-end hook result typing, and SpeechGroupSpanData slots optimization. The release also bumps the Modal sandbox extra to 1.4.3 and adds comprehensive documentation improvements including a MongoDB session memory example, docstring corrections, and external tracing processor updates. Test coverage enhancements were made for streaming paths and retry helpers.

Claude Code v2.1.172 introduces hierarchical sub-agent spawning (up to 5 levels deep), improves AWS region configuration handling for Bedrock, and adds a marketplace search feature. The release includes numerous bug fixes addressing context management, session stability, model selection, and UI responsiveness issues. Performance optimizations reduce CPU usage and improve message handling in long conversations.

langchain-anthropic version 1.4.5 has been released with several improvements and fixes. Key changes include support for content block tokens in callbacks, refreshed model profile data, and a hotfix for OpenAI minimum core dependency. The release also includes test improvements to disable pytest-benchmark under xdist to eliminate warnings. This is a maintenance release focused on stability and compatibility improvements.

LangChain Classic version 1.0.8 is a maintenance release that includes dependency updates, bug fixes, and improvements to core functionality. Key changes involve bumping multiple dependencies (pyarrow, starlette, aiohttp, langsmith, urllib3, and others), fixing content block token support in callbacks, and improving type safety in Runnable operations. The release also includes infrastructure improvements and test enhancements to address warnings and ensure stability.

langchain-core version 1.4.4 is a maintenance release that includes bug fixes, typing improvements, and documentation updates. Key improvements address content block tokens in callbacks, multimodal handling in buffer strings, and tool error handling. The release also enhances type safety across Runnable operations and fixes generic typing issues.

langchain-mistralai version 1.1.5 release includes support for content block tokens in callbacks, stability improvements for integration tests with rate limiting and retries, and multiple dependency updates. Key updates include bumps to langsmith (0.8.5), langchain-core (1.3.3), urllib3 (2.7.0), and idna (3.15). The release focuses on reliability, security patches, and infrastructure hardening for the Mistral AI integration package.

LangChain v1.3.7 release includes style improvements with ruff ARG rules, a new ProviderToolSearchMiddleware feature for enhanced tool searching capabilities, mypy type checking enhancements with warn_return_any activation, and legacy code cleanup marking components for v2.0 removal. This incremental release focuses on code quality, type safety, and middleware functionality improvements.

OpenClaw v2026.6.6-beta.1 is a security-focused release that substantially tightens security boundaries across transcripts, sandbox binds, and multiple integration points while improving Telegram delivery reliability, iMessage recovery, and browser/MCP connectivity. The release also enhances Control UI performance through cached metadata and lazy loading, expands provider support with OpenRouter OAuth and Claude Fable 5, and includes numerous fixes for agent recovery, channel delivery, and authentication. Key improvements span CLI progress reporting, observability, plugin management, memory optimization, and mobile channel support.

LangChain v1.3.5 release includes porting AND-capable trigger conditions to SummarizationMiddleware, adding support for the apply_patch built-in tool in OpenAI integration, and fixing a minimum core dependency issue. The release also includes routine dependency updates for pyarrow and aiohttp to maintain security and compatibility. This is a minor version update with both feature additions and maintenance improvements.

langchain-groq version 1.1.3 is a maintenance release that includes dependency updates, bug fixes, and feature enhancements. Key changes include strict mode support for Groq, content-block-centric streaming improvements, model profile data refresh, and security updates (CVE-2026-4539 pygments fix). The release also standardizes testing infrastructure, updates core dependencies, and improves token extraction logic for usage metadata.

LangChain version 1.3.6 has been released with a focus on maintaining backward compatibility. The primary change in this release is a fix that preserves summarization trigger compatibility, ensuring existing implementations continue to work without breaking changes. This incremental update addresses a specific compatibility issue identified since version 1.3.5.

This cookbook demonstrates two async multi-agent orchestration patterns using Claude: a fixed N-agent team with peer messaging through a shared hub, and dynamically spawned async subagents. Built on the Anthropic Python SDK and asyncio, it provides bare messaging and lifecycle mechanics without domain-specific tasks, allowing developers to see exactly which tools fire and in what order before adding their own tools and logic. The patterns run on any API key in under thirty seconds and include a message hub, messaging tools (send_message and wait_for_message), and a base agent loop that handles tool dispatching and inbox management.

Claude Fable 5 includes safety classifiers that block requests in cybersecurity, biology, and reasoning extraction areas. Anthropic provides server-side and client-side fallback mechanisms to automatically retry blocked requests with Claude Opus 4.8, along with new billing changes that credit fallback token costs. This guide covers classifier block detection, fallback implementation strategies, streaming behavior, and billing adjustments for API customers.

LangChain Core version 1.4.3 has been released with improvements to OpenAI integration. The primary feature addition is support for the `apply_patch` built-in tool, enabling more flexible tool usage within OpenAI-powered agents. This incremental release builds on version 1.4.2 with focused enhancements to the core framework's tooling capabilities.

LangChain OpenAI package version 1.3.0 introduces support for the `apply_patch` built-in tool, updates minimum core dependencies, and refreshes model profile data. This release includes bug fixes, test improvements, and maintenance updates to ensure compatibility and stability. The update builds on version 1.2.2 with enhanced tooling capabilities and improved test infrastructure.

langchain-perplexity version 1.4.0 introduces tool binding capabilities and Responses-API tool round-trip functionality, enabling more sophisticated agent interactions. This release builds on version 1.3.2 with enhanced tool integration features. A hotfix for OpenAI's minimum core dependency is also included to ensure compatibility.

crewAI v1.14.7a4 introduces runtime migration for @listen/@router decorators to read from FlowDefinition and adds pluggable default backends for memory, knowledge, RAG, and flow components. This release enhances the framework's flexibility by allowing developers to swap backend implementations without code changes. The update includes documentation updates and contributions from three core team members.

Claude Code v2.1.170 introduces Claude Fable 5, a powerful Mythos-class model now available for general use with capabilities exceeding all previously released models. The release also fixes a critical bug where sessions launched from VS Code's integrated terminal failed to save transcripts and weren't resumable. Users should update to access Fable 5's enhanced capabilities and benefit from the session persistence fix.

OpenClaw v2026.6.5 is a maintenance release focusing on stability, security, and provider integrations. Key improvements include QQBot reasoning stripping to prevent content leakage, MCP tool result coercion to avoid API errors, Anthropic extended-thinking recovery mechanisms, and Parallel web-search provider bundling. The release also enhances Matrix voice/thread handling, improves auth durability with SQLite storage, and strengthens security around tool access and provider catalogs.

OpenClaw v2026.6.5-beta.6 introduces critical fixes for QQBot reasoning stripping, MCP tool result coercion to prevent API errors, improved Anthropic extended-thinking recovery, Parallel web-search provider bundling, and enhanced durability for auth/plugin state via SQLite. The release strengthens security around MCP leases, prompt-cache tool names, and HTTP tools while improving macOS node stability and upgrade safety across multiple platforms.