v1.14.0-prerelease.2

What's New

You can now create, pause, resume, and upgrade Atlas clusters directly through the AI assistant, as well as load sample data into a cluster with a dedicated tool. Atlas Stream Processing is now supported with a full set of tools for managing streaming pipelines, and search and vector search indexes can be created, listed, and dropped. Tool responses across MongoDB, Atlas, and local tools now include structured content, making results easier for AI assistants to interpret and act on, and sensitive information in logs is redacted by default.

What's Changed

🚀 Features

• feat: enhance validation when creating new db user - MCP-577 by @dudaschar in #1307
• feat: add health-check in documentation and enrich response - MCP-517 by @gusanthiago in #1267

🐛 Bug Fixes

• fix: make requestIdAttr case-insensitive CLOUDP-418875 by @lerouxb in #1301
• fix: enforce redaction of sensitive information in mcp logger by default - MCP-574 by @dudaschar in #1303
• fix: use commit messages for release notes generation MCP-573 by @nirinchev in #1304
• fix: update cluster listing output to include project name inside untrusted tag - MCP-576 by @dudaschar in #1306
• fix: update messages to use "requested collection" and keep collection name along with other parsed content - MCP-580 by @dudaschar in #1314
• fix: do not interpolate names in note for StreamsDiscoverTool output - MCP-578 by @dudaschar in #1308

New Contributors

• @maastha made their first contribution in #1297
• @gribnoysup made their first contribution in #1313
• @gusanthiago made their first contribution in #1267

Full Changelog: v1.14.0-prerelease.1...v1.14.0-prerelease.2