MeshAgent 0.26 brings expansions to web tooling, container infrastructure, and MCP integrations. This release makes it easier to build agents that interact with the web, run code in isolated environments, and connect to external services with better observability and control.
MeshAgent 0.26 expands support for Anthropic's web search and web fetch toolkits, giving your agents powerful capabilities to retrieve and process web content.The new web fetch toolkit goes beyond simple HTML retrieval and now returns structured responses in multiple formats (text, JSON, or file) and automatically converts HTML to Markdown for easier agent processing. This means your agents can search the web, fetch content, and work with it in the format that makes the most sense for their task.
The new container-based shell tool enables you to run agent generated code safely by executing commands in isolated containers with configurable images, mounts, and environment variables. You can now:
1. Specify images for different execution environments
2. Mount specific directories for file access
3. Set environment variables per execution
4. Run code with better isolation and reproducibility. Combined with CLI caching for the shell tool, this makes iterative development faster while maintaining security.
The Model Context Protocol (MCP) bridge gets significant upgrades in 0.26:
1. Streamable HTTP connections for better performance with long-running MCP operations
2. Custom headers and secret-backed headers for authenticated MCP services
3. OAuth2 secrets now accept both text and base64 formats
4. Identity-scoped secrets for clear credential management
These improvements make it easier to integrate third-party MCP services that require authentication or have specific connection requirements.
MeshAgent now includes built-in webhook validation and routing for popular platforms: GitHub, Shopify, Slack, and Twilio. You can wire webhooks directly to queues, making it simple to build agents that respond to external events.
Thread History Improvements
Thread history now enforces a context window and adds helper tools to search, count, and fetch message ranges. This makes it easier to work with long conversations without exceeding model limits.
Tool Decorator Enhancements
The @tool decorator now supports bound instance and class methods, making it simpler to expose agent methods as tools.
Scheduled Tasks View
MeshAgent Studio now includes a scheduled tasks view, making it easy to see and manage periodic agent jobs.
File Preview Improvements
The file preview system got a major upgrade:
1. UTF-8 text loading (enables proper display of special characters in .txt files)
2. Support for BMP, GIF, and SVG image formats
3. Universal image handling that works across different file types
4. Automatic README rendering as Markdown
Join our Discord Server to stay up to date with MeshAgent releases, see full release notes, ask questions, and share feedback with our team.
Check out the MeshAgent documentation to start building today!
MeshAgent Studio, SDK, and Server give you everything to build, test, and deploy agentic applications, from development to production.
