From 133c7015bec042a0eb1af4114541555fb2cdebf2 Mon Sep 17 00:00:00 2001 From: micheleRP Date: Sun, 7 Jun 2026 09:52:08 -0600 Subject: [PATCH 1/2] docs: correct MCP server and AI agent config framing in overview Remove inaccurate "define tools in YAML" / "define behaviors in YAML" claims. MCP server tools are discovered automatically after picking a managed type or registering a self-managed server; agents are configured declaratively (system prompt, model, tools), not authored in YAML. Co-Authored-By: Claude Opus 4.8 (1M context) --- modules/get-started/pages/adp-overview.adoc | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/modules/get-started/pages/adp-overview.adoc b/modules/get-started/pages/adp-overview.adoc index 886c943..4feb4ba 100644 --- a/modules/get-started/pages/adp-overview.adoc +++ b/modules/get-started/pages/adp-overview.adoc @@ -34,7 +34,7 @@ For more information, see xref:gateway:overview.adoc[How AI Gateway Works]. glossterm:MCP server[,MCP servers] translate agent intent into connections to databases, queues, HRIS, CRMs, and other business systems. They are the simplest way to give agents context and capabilities without writing glue code. -MCP servers are lightweight, support OIDC-based authentication, and enforce deterministic policies at the tool level. You define tools in YAML, and policy enforcement programmatically prevents prompt injection, SQL injection, and other agent-based attacks. +MCP servers are lightweight, support OIDC-based authentication, and enforce deterministic policies at the tool level. You pick a managed type from a marketplace and configure it with your credentials, or register a server you already run, and the server's tools are discovered automatically. Policy enforcement programmatically prevents prompt injection, SQL injection, and other agent-based attacks. With real-time debugging capabilities, you reduce integration time while getting enterprise-grade security. You can reuse your existing infrastructure and data sources rather than building new integrations from scratch. @@ -42,9 +42,9 @@ For more information, see xref:connect:mcp-overview.adoc[]. == AI agents -With Redpanda AI agents, you declare the agent behavior you want and Redpanda handles execution and orchestration. Instead of writing Python or JavaScript, you define behaviors in YAML. You can orchestrate multiple specialized glossterm:subagent[,sub-agents], or bring your own frameworks like LangChain or LlamaIndex. +With Redpanda AI agents, you declare the agent behavior you want and Redpanda handles execution and orchestration. Instead of writing Python or JavaScript, you configure agents declaratively: set a system prompt, choose a model, and attach the tools the agent can use. You can orchestrate multiple specialized glossterm:subagent[,sub-agents], or bring your own frameworks like LangChain or LlamaIndex. -What makes this practical at scale is a broad library of prebuilt connectors with built-in filtering, enrichment, and routing, which give declarative definitions real power. +What makes this practical at scale is a broad library of prebuilt tools and integrations with built-in filtering, enrichment, and routing, which give declarative definitions real power. The result is faster time-to-production, lower maintenance (declarative definitions instead of imperative code), and organizational consistency across teams. From 88cb72799f6468367f44dfeab8d6fced828d278d Mon Sep 17 00:00:00 2001 From: micheleRP Date: Mon, 8 Jun 2026 08:27:08 -0600 Subject: [PATCH 2/2] docs: align MCP/agent tool framing across overview and connect pages Incorporate PR review: fix the cross-page inconsistency where agents-overview.adoc still presented agent tools as "prebuilt Redpanda Connect connectors" after the overview switched to neutral "prebuilt tools and integrations" wording. Extend the fix to the MCP server landing pages for consistent vocabulary. - adp-overview.adoc: "register a server you already run" -> "register a self-managed server"; "a marketplace" -> "the marketplace" - agents-overview.adoc: drop "prebuilt connectors" agent-tool framing; tools are sourced through MCP servers; keep Redpanda Connect as the named data/streaming engine and the bidirectional pipeline-agent integration - mcp-servers.adoc, connect/index.adoc: "managed connector" -> "managed type", matching the canonical taxonomy in mcp-overview.adoc Co-Authored-By: Claude Opus 4.8 (1M context) --- modules/connect/pages/agents-overview.adoc | 8 ++++---- modules/connect/pages/index.adoc | 2 +- modules/connect/pages/mcp-servers.adoc | 4 ++-- modules/get-started/pages/adp-overview.adoc | 2 +- 4 files changed, 8 insertions(+), 8 deletions(-) diff --git a/modules/connect/pages/agents-overview.adoc b/modules/connect/pages/agents-overview.adoc index fdd70ae..ce2f771 100644 --- a/modules/connect/pages/agents-overview.adoc +++ b/modules/connect/pages/agents-overview.adoc @@ -1,13 +1,13 @@ = How Agents Work :page-aliases: cloud-data-platform:ai-agents:agents/overview.adoc -:description: Learn how ADP agents use a declarative approach backed by a broad library of prebuilt connectors to replace custom agent code. +:description: Learn how ADP agents use a declarative approach backed by a broad library of prebuilt tools and integrations to replace custom agent code. :page-topic-type: overview :personas: agent_builder, platform_engineer :learning-objective-1: Describe what AI agents are and their essential components :learning-objective-2: Explain how ADP streaming infrastructure benefits agent architectures :learning-objective-3: Identify use cases where ADP agents provide value -Redpanda ADP AI agents take a declarative approach: instead of writing Python or JavaScript agent code, you declare the behavior you want by selecting an LLM, writing a system prompt, and connecting tools drawn from a broad library of prebuilt Redpanda Connect connectors. The framework handles execution, tool orchestration, and scaling, backed by real-time streaming infrastructure and built-in filtering and data enrichment. +Redpanda ADP AI agents take a declarative approach: instead of writing Python or JavaScript agent code, you declare the behavior you want by selecting an LLM, writing a system prompt, and attaching prebuilt tools and integrations. The framework handles execution, tool orchestration, and scaling, backed by real-time streaming infrastructure and built-in filtering and data enrichment. After reading this page, you will be able to: @@ -17,7 +17,7 @@ After reading this page, you will be able to: == What is an AI agent? -An AI agent is a system built around a glossterm:large language model (LLM)[] that interprets user intent, selects the right tools, and chains multiple steps into a workflow. In ADP, agents are declarative: you configure what the agent should do (its role, constraints, and available tools) rather than writing imperative agent code. This is possible because Redpanda Connect provides the connectors and robust data processing capabilities that the framework orchestrates for you. +An AI agent is a system built around a glossterm:large language model (LLM)[] that interprets user intent, selects the right tools, and chains multiple steps into a workflow. In ADP, agents are declarative: you configure what the agent should do (its role, constraints, and available tools) rather than writing imperative agent code. Agents reach tools through MCP servers, and Redpanda Connect provides the real-time streaming and data-processing capabilities the framework orchestrates for you. == How agents work @@ -36,7 +36,7 @@ For a deeper understanding of how agents execute, manage context, and maintain s == Key benefits -A declarative approach means you configure agent behavior instead of coding it, with access to Redpanda Connect's library of prebuilt connectors for data sources, APIs, and services. Real-time streaming data ensures agents access live events instead of batch snapshots. xref:connect:mcp-overview.adoc[Remote MCP] support enables standardized tool access. Managed infrastructure handles deployment, scaling, and security for you. Low-latency execution means tools run close to your data. Integrated secrets management securely stores API keys and credentials. +A declarative approach means you configure agent behavior instead of coding it, with access to a broad library of prebuilt tools and integrations for data sources, APIs, and services. Real-time streaming data ensures agents access live events instead of batch snapshots. xref:connect:mcp-overview.adoc[Remote MCP] support enables standardized tool access. Managed infrastructure handles deployment, scaling, and security for you. Low-latency execution means tools run close to your data. Integrated secrets management securely stores API keys and credentials. == Use cases diff --git a/modules/connect/pages/index.adoc b/modules/connect/pages/index.adoc index 5c14835..067b28b 100644 --- a/modules/connect/pages/index.adoc +++ b/modules/connect/pages/index.adoc @@ -2,4 +2,4 @@ :description: Connect agents, MCP servers, and the apps and data sources behind them. :page-layout: index -Bring your agents, tools, and data into the Agentic Data Plane. Build agents, connect MCP servers and managed connectors, set up authentication, and integrate with your existing stack. +Bring your agents, tools, and data into the Agentic Data Plane. Build agents, connect MCP servers, set up authentication, and integrate with your existing stack. diff --git a/modules/connect/pages/mcp-servers.adoc b/modules/connect/pages/mcp-servers.adoc index 49fcef4..900e2d3 100644 --- a/modules/connect/pages/mcp-servers.adoc +++ b/modules/connect/pages/mcp-servers.adoc @@ -1,5 +1,5 @@ = Connect Tools with MCP Servers -:description: Learn how MCP servers expose tools and data to agents, then plug in managed connectors or build, register, and test your own. +:description: Learn how MCP servers expose tools and data to agents, then plug in a managed type or build, register, and test your own. :page-layout: index -MCP servers expose tools and data to your agents. Plug in a managed connector from the catalog, or build, register, and test a server of your own. +MCP servers expose tools and data to your agents. Plug in a managed type from the catalog, or build, register, and test a server of your own. diff --git a/modules/get-started/pages/adp-overview.adoc b/modules/get-started/pages/adp-overview.adoc index 4feb4ba..388bbab 100644 --- a/modules/get-started/pages/adp-overview.adoc +++ b/modules/get-started/pages/adp-overview.adoc @@ -34,7 +34,7 @@ For more information, see xref:gateway:overview.adoc[How AI Gateway Works]. glossterm:MCP server[,MCP servers] translate agent intent into connections to databases, queues, HRIS, CRMs, and other business systems. They are the simplest way to give agents context and capabilities without writing glue code. -MCP servers are lightweight, support OIDC-based authentication, and enforce deterministic policies at the tool level. You pick a managed type from a marketplace and configure it with your credentials, or register a server you already run, and the server's tools are discovered automatically. Policy enforcement programmatically prevents prompt injection, SQL injection, and other agent-based attacks. +MCP servers are lightweight, support OIDC-based authentication, and enforce deterministic policies at the tool level. You pick a managed type from the marketplace and configure it with your credentials, or register a self-managed server. Policy enforcement programmatically prevents prompt injection, SQL injection, and other agent-based attacks. With real-time debugging capabilities, you reduce integration time while getting enterprise-grade security. You can reuse your existing infrastructure and data sources rather than building new integrations from scratch.