--- title: How we brought agentic workflows to Cloud SIEM with the... description: Datadog's security engineering team shares how they built MCP tools for Cloud SIEM, covering three core challenges: scoping tools around real user behavior... canonical: https://daily.dev/posts/how-we-brought-agentic-workflows-to-cloud-siem-with-the-datadog-mcp-server-giriy3dir twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: How we brought agentic workflows to Cloud SIEM with the Datadog MCP Server | daily.dev og:description: Datadog's security engineering team shares how they built MCP tools for Cloud SIEM, covering three core challenges: scoping tools around real user behavior... og:url: https://daily.dev/posts/how-we-brought-agentic-workflows-to-cloud-siem-with-the-datadog-mcp-server-giriy3dir og:image: https://api.daily.dev/og/posts/gIrIy3DiR.png og:image:alt: How we brought agentic workflows to Cloud SIEM with the Datadog MCP Server 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. # How we brought agentic workflows to Cloud SIEM with the Datadog MCP Server **[Datadog](https://daily.dev/sources/datadog)** · 10 min read · 0 upvotes · 0 comments ## Summary Datadog's security engineering team shares how they built MCP tools for Cloud SIEM, covering three core challenges: scoping tools around real user behavior (using API call patterns and RUM data), managing a shared context window across a growing multi-team toolset via progressive disclosure and query-based triage, and testing non-deterministic agent behavior with a custom eval framework integrated into CI. Key findings include that 44% of user messages involved detection rule authoring, 25% of customers triaged signals in bulk, and redesigning the bulk triage tool to accept search queries instead of signal ID lists reduced triage time from 2+ minutes for 50 signals to ~1 minute for hundreds. A lightweight governance committee enforces minimum eval scores before tools reach customers without slowing iteration. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How can I reduce context window usage when building MCP tools that serve large schemas to an AI agent? Use progressive disclosure by breaking the schema into named sections (top-level fields, query syntax, options, examples) and letting the tool accept filters for the specific rule type or detection method needed. In practice, filtering to a single rule type cut token usage by about 15%, and filtering further by detection method brought the reduction to 41-47%. _Developers designing token-efficient agent tools can find similar architecture breakdowns on daily.dev._ ### How do you test MCP tools when AI agent behavior is non-deterministic? Build a custom eval framework integrated into CI that runs automatically whenever the toolset changes, checking whether the agent called the right tools, avoided the wrong ones, used a sensible number of calls, passed correct arguments, and followed a reasonable trajectory. Running each scenario multiple times with stubbed tool responses averages out noise while keeping costs down, with regressions over 5% flagged and a baseline success target around 80%. _Teams building reliable agent tooling can track eval and testing approaches like this on daily.dev._ ### How many signals can be triaged in a single call with an MCP tool, and how does that compare to the Cloud SIEM UI limit? An MCP triage tool can handle up to 500 signals in a single call, versus the Cloud SIEM UI cap of 50 signals at a time. The redesign resolves signal IDs from a search query rather than passing IDs directly into context, keeping input token cost flat regardless of match count and cutting update time for hundreds of signals to just over a minute, down from at least 2 minutes for 50 signals previously. _Security engineers scaling agent-driven triage workflows can follow tooling patterns like this on daily.dev._ ## Similar posts on daily.dev - [Smarter SIEM starts here: Context, speed, and the power of MCP](https://daily.dev/posts/smarter-siem-starts-here-context-speed-and-the-power-of-mcp-uucvj4th4) · Coralogix · 0 upvotes · 0 comments - [Four ways engineering teams use the Datadog MCP Server to power AI agents](https://daily.dev/posts/four-ways-engineering-teams-use-the-datadog-mcp-server-to-power-ai-agents-kxxssk2xn) · Datadog · 1 upvotes · 0 comments - [How to Investigate a Production Incident Using an AI Agent \(AppSignal MCP\)](https://daily.dev/posts/how-to-investigate-a-production-incident-using-an-ai-agent-appsignal-mcp--5snliyioe) · AppSignal · 0 upvotes · 0 comments - [Build a Multi-Agent GTM Intelligence System](https://daily.dev/posts/build-a-multi-agent-gtm-intelligence-system-2bebbeora) · Daily Dose of Data Science \| Avi Chawla \| Substack · 0 upvotes · 0 comments --- Tags: [#devops](https://daily.dev/tags/devops), [#llm](https://daily.dev/tags/llm), [#ai-agents](https://daily.dev/tags/ai-agents), [#mcp](https://daily.dev/tags/mcp) [View this post on daily.dev](https://daily.dev/posts/how-we-brought-agentic-workflows-to-cloud-siem-with-the-datadog-mcp-server-giriy3dir) ```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":"How we brought agentic workflows to Cloud SIEM with the Datadog MCP Server","url":"https://daily.dev/posts/how-we-brought-agentic-workflows-to-cloud-siem-with-the-datadog-mcp-server-giriy3dir","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/how-we-brought-agentic-workflows-to-cloud-siem-with-the-datadog-mcp-server-giriy3dir"},"datePublished":"2026-07-17T14:55:03.880Z","dateModified":"2026-09-14T07:26:12.636Z","description":"Datadog's security engineering team shares how they built MCP tools for Cloud SIEM, covering three core challenges: scoping tools around real user behavior...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/299f40f28ef2a652950157dcbce7e24d?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/299f40f28ef2a652950157dcbce7e24d?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Datadog","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":"Datadog","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/8f9d7e3e4bd64bd7b9db9b2ea83e0c6a","url":"https://daily.dev/sources/datadog"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/how-we-brought-agentic-workflows-to-cloud-siem-with-the-datadog-mcp-server-giriy3dir","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"devops,llm,ai-agents,mcp","timeRequired":"PT10M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Datadog","item":"https://daily.dev/sources/datadog"},{"@type":"ListItem","position":3,"name":"How we brought agentic workflows to Cloud SIEM with the Datadog MCP Server"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/how-we-brought-agentic-workflows-to-cloud-siem-with-the-datadog-mcp-server-giriy3dir#faq","mainEntity":[{"@type":"Question","name":"How can I reduce context window usage when building MCP tools that serve large schemas to an AI agent?","acceptedAnswer":{"@type":"Answer","text":"Use progressive disclosure by breaking the schema into named sections (top-level fields, query syntax, options, examples) and letting the tool accept filters for the specific rule type or detection method needed. In practice, filtering to a single rule type cut token usage by about 15%, and filtering further by detection method brought the reduction to 41-47%. Developers designing token-efficient agent tools can find similar architecture breakdowns on daily.dev."}},{"@type":"Question","name":"How do you test MCP tools when AI agent behavior is non-deterministic?","acceptedAnswer":{"@type":"Answer","text":"Build a custom eval framework integrated into CI that runs automatically whenever the toolset changes, checking whether the agent called the right tools, avoided the wrong ones, used a sensible number of calls, passed correct arguments, and followed a reasonable trajectory. Running each scenario multiple times with stubbed tool responses averages out noise while keeping costs down, with regressions over 5% flagged and a baseline success target around 80%. Teams building reliable agent tooling can track eval and testing approaches like this on daily.dev."}},{"@type":"Question","name":"How many signals can be triaged in a single call with an MCP tool, and how does that compare to the Cloud SIEM UI limit?","acceptedAnswer":{"@type":"Answer","text":"An MCP triage tool can handle up to 500 signals in a single call, versus the Cloud SIEM UI cap of 50 signals at a time. The redesign resolves signal IDs from a search query rather than passing IDs directly into context, keeping input token cost flat regardless of match count and cutting update time for hundreds of signals to just over a minute, down from at least 2 minutes for 50 signals previously. Security engineers scaling agent-driven triage workflows can follow tooling patterns like this on daily.dev."}}]} ```