AI Tool Updates - June 9, 2026

Week of: June 9, 2026

---

Overview

This week's updates span bug-fix releases and new features across n8n, Claude Code, Ollama, LangChain, and the Vercel AI SDK. Notable highlights include Claude Code's new safe-mode troubleshooting flag and directory-switching command, Ollama's launch of Hermes Desktop, and GitHub's addition of periodic code scanning for inactive repositories.

Stories

1. Claude Code v2.1.169 Adds Safe Mode and Directory-Switching Command

Source: Claude Code Link: https://github.com/anthropics/claude-code/releases/tag/v2.1.169

Claude Code v2.1.169 introduces a --safe-mode flag (and CLAUDE_CODE_SAFE_MODE environment variable) that starts the agent with all customizations—CLAUDE.md, plugins, skills, hooks, and MCP servers—disabled for troubleshooting. A new /cd command lets users move a session to a new working directory without breaking the prompt cache mid-session.

The release also adds a disableBundledSkills setting and corresponding environment variable to hide bundled skills, workflows, and built-in slash commands from the model. Input navigation was fixed so Up/Down arrows move through each visual row of a long input line instead of jumping to command history.

Impact Analysis: The safe-mode flag gives AI builders a clean troubleshooting path when customizations cause unexpected behavior, and the /cd command preserves prompt cache across directory changes, reducing latency in multi-directory workflows.

2. n8n Releases v2.26.0 Beta with Security and Credential Management Fixes

Source: n8n Link: https://github.com/n8n-io/n8n/releases/tag/beta

n8n's beta channel advanced to v2.26.0, featuring bug fixes including enforcement of a redaction floor when creating workflows via the API, proper handling of all binary data modes in the AWS Rekognition Node, and support for exporting workflows with AI Gateway-managed credentials. The release also patches CVEs in hono and postcss identified by Trivy nightly scans.

Additional core fixes deduplicate getConnectedNodes results, delete orphaned per-user entries on unshare and membership changes, and bump dependencies including qs, tmp, and uuid.

Impact Analysis: The AI Gateway credential export fix and redaction enforcement are directly relevant for teams building AI workflows with n8n that require secure credential management and compliance.

3. Ollama v0.30.7 Launches Hermes Desktop Interface

Source: Ollama Link: https://github.com/ollama/ollama/releases/tag/v0.30.7

Ollama v0.30.7 makes Hermes Desktop available via ollama launch hermes-desktop, providing a native desktop interface for managing conversations, integrations, and messaging apps alongside the Hermes agent. The release also aligns the OpenAI-compatible API models list with available model tags and adds native Windows configuration path support.

Impact Analysis: This release lowers the barrier for running agent-based workflows locally with a visual interface, making Ollama a more complete platform for prototyping AI toolchains on desktop.

4. GitHub Adds Periodic Code Scanning for Inactive Repositories

Source: GitHub Changelog Link: https://github.blog/changelog/2026-06-09-periodic-code-scanning-of-inactive-repositories

GitHub code scanning now supports scheduled security scans of repositories that have had no pushes or pull requests for six months or more. Organizations can use this feature to maintain continuous security coverage across dormant codebases.

Impact Analysis: For teams managing AI tool repositories or SDKs that may become inactive, this automated scanning ensures security vulnerabilities don't go undetected in archived or low-activity projects.

5. Vercel AI SDK Ships Canary Updates with Workflow Rename and TUI Agent Support

Source: Vercel AI SDK Link: https://github.com/vercel/ai/releases/tag/%40ai-sdk%2Fworkflow%401.0.0-canary.83

The Vercel AI SDK published multiple canary releases across its packages. The @ai-sdk/workflow canary renamed onStepFinish to onStepEnd, while @ai-sdk/tui reached canary.1 with a new "agent tui" feature. The core ai package advanced to v7.0.0-canary.166, with corresponding updates to React, Vue, Svelte, and RSC bindings.

Impact Analysis: The workflow step callback rename signals API stabilization ahead of a stable release, and the new TUI agent support opens the door for terminal-based AI agent interfaces built on the SDK.

6. LangChain Core v1.4.2 Released with Minor Update

Source: LangChain Link: https://github.com/langchain-ai/langchain/releases/tag/langchain-core%3D%3D1.4.2

LangChain core advanced to v1.4.2, adding a dict() method to the core package. The release is a minor patch following v1.4.1.

Impact Analysis: The new dict() method provides a convenience utility for LangChain developers working with structured data serialization in AI pipeline components.

Source Links

• Claude Code - Claude Code v2.1.169 Adds Safe Mode and Directory-Switching Command
• n8n - n8n Releases v2.26.0 Beta with Security and Credential Management Fixes
• Ollama - Ollama v0.30.7 Launches Hermes Desktop Interface
• GitHub Changelog - GitHub Adds Periodic Code Scanning for Inactive Repositories
• Vercel AI SDK - Vercel AI SDK Ships Canary Updates with Workflow Rename and TUI Agent Support
• LangChain - LangChain Core v1.4.2 Released with Minor Update