Tool Use

Claude Code v2.1.178 introduces enhanced permission rules with parameter matching syntax, improved nested `.claude/` directory handling for skills and workflows, and refined auto mode with pre-launch subagent evaluation. The release includes better error messaging for Remote Control, improved `/doctor` output, and numerous bug fixes addressing subagent transcript display, authentication issues, and vim mode undo behavior.

Google ADK Python v1.35.1 is a bug fix release addressing two critical issues: grounding metadata handling for 3.1 live interactions and streaming frame completion. The release ensures proper metadata transmission at turn boundaries and guarantees final partial frames are correctly yielded in streaming scenarios.

Propolis is an autonomous QA platform that deploys swarms of browser agents to simulate user behavior, identify bugs, and generate e2e tests for web applications. The agents collaboratively explore websites, flag friction points, and propose tests that integrate into CI/CD pipelines. Available at $1000/month with flexible pricing options, it addresses the gap between deterministic testing and real-world usage coverage by treating agents as a canary group for quality assurance.

Mosaic is an agentic video editing platform that uses multimodal AI and a node-based canvas interface to automate video editing workflows. Built by former Tesla engineers, it addresses frustrations with traditional editors by enabling users to create reusable editing agents that can analyze video content and apply intelligent edits through natural language prompts. The platform combines visual intelligence (saliency analysis, object detection, emotion recognition) with a timeline editor and supports export to DaVinci Resolve, Premiere Pro, and Final Cut Pro.

OpenClaw v2026.6.8-beta.1 introduces significant improvements to channel delivery (Telegram rich text, WhatsApp ACP bindings), agent and gateway recovery mechanisms, expanded provider/model support (GLM-5.2, Claude Haiku 4.5), enhanced usage hooks with footer rendering, and UI/mobile stability fixes. The release addresses 50+ issues across messaging, runtime reliability, authentication, memory management, and user interface components, with contributions from 30+ developers.

OpenClaw v2026.6.7-beta.1 delivers significant improvements across channel delivery, provider resilience, security boundaries, and recovery paths. Key enhancements include tighter Slack/Telegram integration, better model handling for Kimi/Mistral/DeepSeek/Anthropic, safer auth contexts, and improved failure recovery for agents and cron jobs. The release also strengthens UI/docs accessibility, Docker bundling, and QA validation with new scorecard taxonomy artifacts.

langchain-openai version 1.3.1 release includes documentation updates, test improvements for deserialization and streaming, package version tracking enhancements, and bug fixes for tool call normalization and structured output handling. Key changes focus on improving type checking, tracing metadata, and test coverage across the LangChain ecosystem. The release maintains compatibility while strengthening internal validation and documentation standards.

langchain-anthropic version 1.4.6 release includes critical fixes for file-search results confinement and tightened Anthropic allowed_prefixes validation. The release also incorporates package version tracking in tracing metadata, upgraded mypy to 2.1, enhanced tool call chunk validation during streaming, and improved test robustness for gateway base URLs. Key improvements focus on security, type safety, and testing reliability across the LangChain-Anthropic integration.

langchain-core version 1.4.7 is released with several maintenance and bug fixes. Key updates include a tornado dependency bump, fixes for Pydantic v1 support in tools/runnable, corrections to package version trace metadata, and documentation style improvements across docstrings. This is a minor patch release focused on stability and compatibility improvements.

LangChain version 1.3.9 has been released with updates to the Anthropic integration (1.4.6) and security improvements. The release includes a fix that confines file-search results and tightens the `allowed_prefixes` configuration for Anthropic, enhancing security and control over file access patterns. This is a minor patch release following version 1.3.8.

OpenClaw v2026.6.6 is a security and stability-focused release that substantially tightens security boundaries across transcripts, sandboxing, and multiple integrations 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 security hardening, messaging platform reliability, performance optimization, and provider compatibility.

Claude Code v2.1.175 introduces the `enforceAvailableModels` managed setting, which strengthens model access control by constraining the Default model selection to the allowlist. When enabled, any Default model that resolves to a disallowed model automatically falls back to the first allowed model, and users cannot override the managed allowlist with broader settings.

LangChain v1.3.8 release includes documentation improvements, type system enhancements, and bug fixes across the monorepo. Key updates involve replacing double backticks in docstrings for consistency, adding overloads to `create_agent`, improving async middleware decorator typing, and tightening structured output model fallbacks. The release also includes mypy version bump to 2.1 and unified type-check configuration across the project.

Claude Code v2.1.174 release includes UI improvements for model selection, fixes for environment variable inheritance in background sessions, corrections to Bedrock GovCloud region handling, and enhanced usage attribution tracking in VSCode. The update addresses multiple edge cases affecting model display, authentication, and performance, particularly for enterprise users with specialized billing configurations.

OpenClaw v2026.6.6-beta.2 is a security-focused beta release that substantially tightens security boundaries across transcripts, sandboxing, and MCP integrations while improving Telegram delivery reliability, iMessage recovery, browser/MCP connectivity, and Control UI startup performance. The release expands provider support with OpenRouter OAuth and Claude Fable 5, introduces observability enhancements for diagnostics, and fixes critical issues in agent session recovery, channel delivery, and authentication. Multiple infrastructure improvements address performance bottlenecks, plugin management, and build reliability.

crewAI v1.14.7 introduces pluggable backends for memory, knowledge, RAG, and flow management, along with a native Snowflake Cortex LLM provider and chat API for conversational flows. The release includes significant bug fixes for checkpoint restoration, telemetry, file input handling, and security vulnerabilities. Performance improvements focus on lazy-loading docling imports to speed up crewai imports, while refactoring decouples conversational logic and splits flow.py into modular components.

Claude Code v2.1.173 addresses two key fixes: normalization of Fable 5 model names that include a [1m] suffix (representing 1M context) by automatically stripping it, and elimination of a false "sandbox dependencies missing" warning on Windows systems when sandbox mode is enabled. These updates improve model name handling and reduce unnecessary startup warnings for Windows users.

langchain-core version 1.4.5 introduces validation for tool call chunks during streaming, fixes an async tracer fallback issue in sync contexts, and tightens structured output model fallbacks. This release improves streaming reliability and error handling for chat models and structured outputs.

langchain-core version 1.4.6 introduces package version tracking in tracing metadata for better observability, fixes normalization of v1 streamed tool calls in OpenAI integration, and upgrades mypy to 2.1 with unified type-checking configuration across the monorepo. This release enhances tracing capabilities, improves tool call handling consistency, and strengthens code quality standards.

langchain-model-profiles version 0.0.6 is a maintenance and feature release that includes validation of tool call chunks during streaming, dependency updates across multiple packages (langsmith, urllib3, langchain-core), security patches (pygments CVE-2026-4539), and infrastructure improvements. The release encompasses 80+ commits with focus on stability, testing enhancements, and schema validation for model profiles. Key additions include new model profile fields (text_inputs, text_outputs), improved streaming support with content-block-centric v2, and various bug fixes across partner integrations.