Workato's Google Workspace MCP uses one tool, 1,000+ AI-reviewed self-updating actions
adamse · x · 2026-10-24
- Workato enterprise architect Josh Cramoy details their Google Workspace MCP server: one tool covering all seven apps (Gmail, Drive, Calendar, Docs, Sheets, Slides, Contacts) via a semantic layer of 1,000+ known-good actions.
- The core problem solved is context cost: traditional setups mount 7 MCP servers with dozens of tool schemas resident in context and billed every turn, plus endless manual maintenance.
- Self-maintaining loop: an offline learner discovers new Google capabilities on its own; a capability ships only after surviving a canary test and an adversarial review where one model argues for it, another against, and a third decides. No engineering needed when Google ships new features.
More from coding & agent
- Factory tested Reflection's Beam early via Together AI on real coding and agentic workflows — togethercompute · 2026-10-24
- Seroter's Reading List: CLIs Save Tokens, Agent Harness Explained, When to Use Diffusion LLMs — rseroter · 2026-10-24
- FrontierCS 2 benchmark makes agents beat original methods from code with the paper's contribution deleted — berkeley_ai · 2026-10-24
- Dev's workflow: pairing Grok bot with open-source Hermes Agent side by side — alexcovo_eth · 2026-10-24
- Anthropic's Claude Duet: a proactive agent that keeps working after you close the app — edwinarbus · 2026-10-24
- Infinity: a Rust agent runtime that fits 75k agents in a Raspberry Pi's RAM — ShadajL · 2026-10-24