Skills & Tools

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.

Claude Code v2.1.129 introduces plugin URL fetching, improved output synchronization, and auto-update capabilities for package managers. The release includes refinements to model discovery, history search behavior, and skill overrides, along with numerous bug fixes addressing UI regressions, cache handling, OAuth issues, and policy enforcement. Key improvements enhance terminal compatibility, conversation management, and enterprise security features.

langchain-fireworks version 1.3.1 is a patch release that addresses three key issues: requiring explicit API key configuration in FireworksEmbeddings, fixing a release versioning issue, and correcting how ToolMessage text content blocks are processed by removing non-wire protocol keys. This update ensures proper authentication handling and improves compatibility with the Fireworks API.

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.

OpenClaw v2026.5.4-beta.1 introduces a bundled file-transfer plugin for binary file operations, enhances Google Meet/Voice Call integration with Twilio's realtime Gemini voice bridge, and delivers significant performance improvements across the gateway startup, plugin loading, and UI rendering. The release includes numerous UX refinements for chat controls, streaming progress visualization, and agent command steering, plus expanded QA/testing capabilities for Discord and Slack integrations.

OpenClaw v2026.5.3 introduces a bundled file-transfer plugin with secure binary file operations, hardens plugin installation workflows, and optimizes Gateway startup performance through lazy-loading. The release improves channel reliability across Discord, WhatsApp, Telegram, and other platforms, adds new agent steering commands, and fixes critical issues in realtime transcription, plugin updates, and systemd secret management.

OpenClaw v2026.5.3-beta.3 introduces a bundled file-transfer plugin with secure binary file operations, hardens plugin installation workflows, and optimizes Gateway performance through lazy-loading. Major improvements include enhanced channel support (WhatsApp, Discord, Telegram), improved agent runtime reliability, and numerous fixes for edge cases in streaming, transcription, and configuration management. The release emphasizes security (default-deny policies, symlink traversal protection), performance (startup optimization, hot-path trimming), and operational robustness (doctor command enhancements, state recovery).

OpenClaw v2026.5.3-beta.2 introduces a bundled file-transfer plugin with secure binary file operations, hardens plugin installation and updates, and improves gateway performance through lazy-loading. The release enhances channel reliability across Discord, WhatsApp, Telegram, and other platforms, while fixing critical issues in agent runtime, memory status reporting, and macOS upgrades. Key improvements focus on security (path policies, symlink traversal), performance (startup optimization), and reliability (streamed replies, session recovery).