Top Story
[Release] anthropics/claude-code v2.1.117: v2.1.117
Claude Code v2.1.117 introduces significant improvements to subagent forking, MCP server handling, model selection persistence, and plugin management. The release includes performance optimizations for startup and file searches, enhanced OpenTelemetry observability, and numerous bug fixes across OAuth, WebFetch, proxy handling, and context window calculations. Key additions include experimental Advisor Tool improvements, native build optimizations using embedded tools, and better error handling for plugin dependencies and marketplace enforcement.
- Enable forked subagents on external builds with `CLAUDE_CODE_FORK_SUBAGENT=1` environment variable
- Agent frontmatter `mcpServers` now load for main-thread sessions via `--agent` flag
- Model selections persist across restarts independently of project pins; startup header indicates source (project or managed-settings)
- +7 more key points...
By Topic
Agent Teams
See all 6 →[Release] crewaiInc/crewAI 1.14.3a2: 1.14.3a2
crewAI v1.14.3a2 introduces support for AWS Bedrock V4 and Daytona sandbox tools while expanding documentation with an AI-native 'Build with AI' page across multiple languages. The release includes critical bug fixes for event propagation, batch initialization, task serialization, and guardrail handling, plus a security update to python-dotenv. This alpha release enhances the platform's capabilities for AI agent development with improved infrastructure support and developer resources.
greysonlalonde
AutoGen Custom Reply Functions 🔥 Control AI Responses Like a Pro (Router + Preprocessing)
This tutorial demonstrates how to use custom reply functions in AutoGen to control AI agent responses with advanced routing and preprocessing techniques. Learn to implement router patterns and preprocessing logic to customize how agents respond to messages, enabling more sophisticated multi-agent workflows. The course includes practical GitHub examples for building production-ready AutoGen applications.
Nidhi Chouhan
สอนใช้ “Manus AI” คนเดียวก็สร้างธุรกิจได้ | TNN Tech Reports Weekly
Manus AI is an emerging technology platform enabling solo entrepreneurs and SME business owners to automate work tasks and scale operations efficiently. The platform allows individuals to create and deploy AI agents that can perform real business functions, reducing the need for large teams. This technology represents a significant opportunity for small business owners to compete with larger enterprises by leveraging AI automation.
TNN
Integrations
[Release] langchain-ai/langchain langchain-openai==1.1.16: langchain-openai==1.1.16
langchain-openai version 1.1.16 has been released with a critical bug fix addressing `prompt_cache_retention` drift issues in streaming operations. This patch resolves compatibility issues that could occur when streaming responses with prompt caching enabled. The fix ensures more robust handling of cache retention parameters during streaming scenarios.
github-actions[bot]
OpenClaw + Google Workspace: The Ultimate AI Setup
This content describes an integration between OpenClaw and Google Workspace that enables AI agents to perform practical work tasks like calendar organization, document drafting, and workspace automation. The setup creates an 'AI employee' that goes beyond conversational AI to actively manage business workflows and productivity tools. This represents a practical implementation of AI agents in real business environments using accessible tools.
Phil Lougher
Security
Skills & Tools
Multi-Agent Systems
Use Cases
Today's Breakdown
Sources
Content Types
Difficulty
Want to save articles & build playbooks?
Create a free account to save articles to your library, build AI-generated implementation playbooks, and get content matched to your stack.
Or just get it by email
No account needed. Unsubscribe anytime.