--- title: Uber Builds MCP Gateway to Govern Access to 800+ Servers... description: Uber built an MCP Gateway to standardize how AI agents discover and invoke internal services, supporting over 800 MCP servers and 5,000 tools. The architecture... canonical: https://daily.dev/posts/uber-builds-mcp-gateway-to-govern-access-to-800-servers-and-5-000-ai-agent-tools-pzxako5ck twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Uber Builds MCP Gateway to Govern Access to 800+ Servers and 5,000+ AI Agent Tools | daily.dev og:description: Uber built an MCP Gateway to standardize how AI agents discover and invoke internal services, supporting over 800 MCP servers and 5,000 tools. The architecture... og:url: https://daily.dev/posts/uber-builds-mcp-gateway-to-govern-access-to-800-servers-and-5-000-ai-agent-tools-pzxako5ck og:image: https://api.daily.dev/og/posts/PZXAKO5CK.png og:image:alt: Uber Builds MCP Gateway to Govern Access to 800+ Servers and 5,000+ AI Agent Tools og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Uber Builds MCP Gateway to Govern Access to 800+ Servers and 5,000+ AI Agent Tools **[InfoQ](https://daily.dev/sources/infoq)** · 3 min read · 0 upvotes · 0 comments ## Summary Uber built an MCP Gateway to standardize how AI agents discover and invoke internal services, supporting over 800 MCP servers and 5,000 tools. The architecture splits a control-plane Registry, which stores tool definitions, schemas, ownership and enablement settings, from a Proxy Gateway that translates MCP requests into HTTP, gRPC or TChannel calls at runtime. Uber's AutoCrawler uses Cadence to scan internal IDL registries and auto-generate MCP schemas from Protocol Buffers and Thrift, with new tools disabled until teams review them. Omni MCP adds incremental tool discovery and Response Projection to avoid loading thousands of tool definitions into an agent's context at once. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How does Uber's MCP Gateway separate tool registration from runtime execution? Uber splits the architecture into an MCP Registry acting as the control plane, which stores tool definitions, schemas, ownership and enablement settings, and a Proxy Gateway that handles runtime execution. The proxy translates MCP requests into HTTP, gRPC, or TChannel calls and converts responses back into MCP-compatible results, without requiring changes to underlying services. _daily.dev surfaces architecture deep dives like this for teams designing their own MCP infrastructure._ ### How does Uber automatically generate MCP tool definitions from existing APIs? Uber's AutoCrawler, built on the Cadence workflow engine, scans internal interface definition language registries and extracts API definitions from Protocol Buffers and Thrift schemas. It generates MCP-compatible schemas and uses a large language model to write tool descriptions. Newly discovered tools stay disabled until the owning team reviews and explicitly enables them. _Follow practical MCP tooling patterns like AutoCrawler on daily.dev as agent infrastructure matures._ ### How do AI agents handle tool discovery when there are thousands of available tools? Uber's Omni MCP supports incremental discovery, letting agents search for servers, find relevant tools, retrieve individual schemas, and invoke only the tools needed, instead of loading every tool definition into context upfront. It pairs this with Response Projection, which lets agents request specific response fields rather than full payloads, reducing context overhead at scale with over 5,000 tools. _daily.dev tracks scaling patterns like incremental tool discovery for engineers building agent context pipelines._ ## Similar posts on daily.dev - [How DoorDash Built a Toolbox for AI Agents](https://daily.dev/posts/how-doordash-built-a-toolbox-for-ai-agents-bufaj6vpo) · ByteByteGo · 7 upvotes · 0 comments - [What Is MCP and Why Does It Need a Gateway? A Practical Guide for AI Engineers](https://daily.dev/posts/what-is-mcp-and-why-does-it-need-a-gateway-a-practical-guide-for-ai-engineers-fhf76gbfk) · DEV · 6 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#mcp](https://daily.dev/tags/mcp), [#microservices](https://daily.dev/tags/microservices), [#uber](https://daily.dev/tags/uber), [#api-gateway](https://daily.dev/tags/api-gateway) [View this post on daily.dev](https://daily.dev/posts/uber-builds-mcp-gateway-to-govern-access-to-800-servers-and-5-000-ai-agent-tools-pzxako5ck) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Uber Builds MCP Gateway to Govern Access to 800+ Servers and 5,000+ AI Agent Tools","url":"https://daily.dev/posts/uber-builds-mcp-gateway-to-govern-access-to-800-servers-and-5-000-ai-agent-tools-pzxako5ck","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/uber-builds-mcp-gateway-to-govern-access-to-800-servers-and-5-000-ai-agent-tools-pzxako5ck"},"datePublished":"2026-11-02T14:03:23.296Z","dateModified":"2026-11-04T08:09:31.647Z","description":"Uber built an MCP Gateway to standardize how AI agents discover and invoke internal services, supporting over 800 MCP servers and 5,000 tools. The architecture...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/994d898a11a19013662811d041db6faf?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/994d898a11a19013662811d041db6faf?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"InfoQ","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"InfoQ","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/afc3bced3e1e4b188dd9127017a60e0c","url":"https://daily.dev/sources/infoq"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/uber-builds-mcp-gateway-to-govern-access-to-800-servers-and-5-000-ai-agent-tools-pzxako5ck","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,mcp,microservices,uber,api-gateway","timeRequired":"PT3M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"InfoQ","item":"https://daily.dev/sources/infoq"},{"@type":"ListItem","position":3,"name":"Uber Builds MCP Gateway to Govern Access to 800+ Servers and 5,000+ AI Agent Tools"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/uber-builds-mcp-gateway-to-govern-access-to-800-servers-and-5-000-ai-agent-tools-pzxako5ck#faq","mainEntity":[{"@type":"Question","name":"How does Uber's MCP Gateway separate tool registration from runtime execution?","acceptedAnswer":{"@type":"Answer","text":"Uber splits the architecture into an MCP Registry acting as the control plane, which stores tool definitions, schemas, ownership and enablement settings, and a Proxy Gateway that handles runtime execution. The proxy translates MCP requests into HTTP, gRPC, or TChannel calls and converts responses back into MCP-compatible results, without requiring changes to underlying services. daily.dev surfaces architecture deep dives like this for teams designing their own MCP infrastructure."}},{"@type":"Question","name":"How does Uber automatically generate MCP tool definitions from existing APIs?","acceptedAnswer":{"@type":"Answer","text":"Uber's AutoCrawler, built on the Cadence workflow engine, scans internal interface definition language registries and extracts API definitions from Protocol Buffers and Thrift schemas. It generates MCP-compatible schemas and uses a large language model to write tool descriptions. Newly discovered tools stay disabled until the owning team reviews and explicitly enables them. Follow practical MCP tooling patterns like AutoCrawler on daily.dev as agent infrastructure matures."}},{"@type":"Question","name":"How do AI agents handle tool discovery when there are thousands of available tools?","acceptedAnswer":{"@type":"Answer","text":"Uber's Omni MCP supports incremental discovery, letting agents search for servers, find relevant tools, retrieve individual schemas, and invoke only the tools needed, instead of loading every tool definition into context upfront. It pairs this with Response Projection, which lets agents request specific response fields rather than full payloads, reducing context overhead at scale with over 5,000 tools. daily.dev tracks scaling patterns like incremental tool discovery for engineers building agent context pipelines."}}]} ```