Suman Bisht

product marketing

<p>Organizations running Red Hat OpenShift have already solved the hardest parts of enterprise infrastructure: RBAC, network isolation, security context constraints, runtime scanning, and audit logging. The next challenge is what happens when developers start running AI coding agents inside that infrastructure: which tools and models agents can reach, what network access they have, and how to produce an audit trail that helps with troubleshooting and auditing.</p>

<p>Coder runs natively on OpenShift and adds a governance layer to the AI development environment. With the latest updates, Coder adds AI governance and agent-based automation (Coder Agents in early access), making it much easier to support enterprise AI workflows without losing oversight.</p>

<p>This post covers how Coder and Red Hat together deliver a self-hosted, governed platform for running AI agents.</p>

<p><img src="https://www.datocms-assets.com/19109/1777064510-blog_coder-redhat-summit.png" alt=""></p>

<h2>Coder AI Governance 101</h2>

<p><a href="https://coder.com/solutions/ai-governance"><em>Coder AI Governance</em></a> gives platform and security teams visibility into how AI tools and agents are used across development environments. It provides centralized control over model access, network behavior, and audit logging. It has two key capabilities: </p>

<p><a href="https://coder.com/docs/ai-coder/ai-gateway"><em>AI Gateway</em></a> is a centralized LLM gateway that sits between AI tools such as Claude Code,  Codex, Aider, custom agents, and upstream model providers like Anthropic, OpenAI, AWS Bedrock, or self-hosted models on OpenShift AI.</p>

<ul>
<li>Centralized authentication: Manage access to AI models, reducing security risk and operational overhead. Eliminate LLM credential sprawl and brittle key rotation while also supporting BYOK with authentication and observability.</li>
<li>Full prompt audit trails: Groups intercepted requests into sessions and threads to show the causal relationships between human prompts and agent actions, providing auditors with clear provenance over who initiated what, when, and why.</li>
<li>Cost visibility: Routes all the traffic through the governance layer, enabling per-user token attribution and spend visibility. Applies quotas and workspace limits to control and optimize token spend.</li>
<li>Provider flexibility: route to Anthropic, OpenAI, Bedrock, or self-hosted models on OpenShift AI.</li>
</ul>

<p><a href="https://coder.com/docs/ai-coder/agent-firewall"><em>Agent Firewall</em></a> is a process-level firewall that restricts and audits what AI agents can access over the network. Unlike traditional network policies that apply to the entire pod, Agent Firewall operates at the process level, giving platform teams granular control over agent behavior without affecting developer workflows.</p>

<ul>
<li>Domain allowlisting: define exactly which domains and HTTP methods an agent can reach. Everything else is blocked.</li>
<li>Centralized audit logs: every agent HTTP request (allowed or denied) is streamed to the Coder control plane and queryable via Grafana Loki and ingestible by enterprise SIEMs such as Splunk.</li>
<li>Landjail mode: Agent Firewall runs entirely without elevated permissions or added capabilities, so it works out-of-the-box on hardened OpenShift clusters. A privileged mode exists for non-OpenShift environments where capability-based enforcement is preferred.</li>
</ul>

<p>Agent Firewall enforces allowlist-only network policies. Admins define which domains agents can reach. All network requests are logged and auditable, helping reduce risk of unauthorized access, package misuse, and data exposure. Security teams can audit what happened and why.</p>

<p>Here is an example Agent Firewall rules file for a  DoD / public-sector deployment:</p>

<pre><code class="yaml">rules:
  - domain: github.example.mil
    methods: [GET, POST, PUT, PATCH]
  - domain: registry.example.mil
    methods: [GET]
  - domain: llm.ocp-ai.example.mil
    methods: [POST]
  # Everything else: DENIED by default
</code></pre>

<h3><strong>Two ways to run AI agents with Coder</strong></h3>

<ul>
<li>Third-party agents: Claude Code, Codex, etc., installed in the Coder workspace itself, governed by AI Gateway and Agent Firewall.</li>
<li>Coder Agents: Coder’s native agent capability (Early Access). The agent loop runs in the control plane; developers interact via a built-in chat UI or API for autonomous agentic workflows. Workspaces are only spun up when the agent needs to make a code change.</li>
</ul>

<p><strong>Note:</strong> Teams can choose to run AI agents both ways mentioned above. Admins can standardize on Coder Agents as the default experience for most developers, while still enabling Claude Code or Codex inside workspaces for specific workflows. Same Coder deployment, same templates, same identity and access controls.</p>

<p><a href="https://coder.com/docs/ai-coder/agents"><em>Coder Agents</em></a> is a self-hosted AI coding agent that runs the agent loop in the Coder control plane rather than inside the workspace. This architecture has specific security and operational properties:</p>

<ul>
<li>Self-hosted model support with consistent governance: Coder Agents works with Anthropic, OpenAI, Google, Azure, AWS Bedrock, or any OpenAI-compatible endpoint including self-hosted models on RHOAI. Sub-agents spawned for parallel work each run in their own context window and are all governed by the same firewall as the root agent.</li>
<li>No LLM API Keys in workspaces: The control plane makes all outbound requests to model providers. Provider credentials are managed centrally, separate from the workspace environment where source code and developer tooling live.</li>
<li>No privilege escalation: the agent operates with the exact same permissions as the user who submitted the prompt. No shared service accounts.</li>
<li>User identity always attached: every action the agent takes (PRs, code pushed, commands run) is tied to the user who submitted the prompt.</li>
<li>Workspace isolation preserved: the agent can only access workspaces owned by the submitting user.</li>
</ul>

<p>This differs from third-party agents, which execute inside the workspace and require additional controls for network access and credentials.</p>

<h2>The Red Hat + Coder stack</h2>

<p>The integration between Red Hat and Coder spans infrastructure, model serving, and governance:</p>

<table><thead>
<tr>
<th><strong>Red Hat Provides</strong></th>
<th><strong>Coder Provides</strong></th>
<th><strong>Together</strong></th>
</tr>
</thead><tbody>
<tr>
<td>OpenShift Container Platform (RHOCP): Enterprise standard Kubernetes based platform</td>
<td>Isolated Workspaces with AI Governance Add-On (AI Gateway, Agent Firewall)</td>
<td>Governed AI agent workspaces on hardened infrastructure, auditable end-to-end</td>
</tr>
<tr>
<td>OpenShift AI (RHOAI): LLM model serving via vLLM, llm-d distributed inference</td>
<td>AI Gateway: routes developer/agent traffic to RHOAI inference endpoints</td>
<td>Self-hosted AI coding pipeline: models served on RHOAI, governed by Coder, no data leaves the perimeter</td>
</tr>
<tr>
<td>RHEL UBI base images: hardened container foundations suitable for regulated environments</td>
<td>Workspace templates: Terraform-defined environments on OCP namespaces</td>
<td>Compliant-by-default developer and agent environments that pass ATO review</td>
</tr>
<tr>
<td>Advanced Cluster Security (RHACS): runtime container security, vulnerability scanning</td>
<td>Agent Firewall: process-level domain allowlists</td>
<td>Defense in depth: container-level security from ACS + process-level agent boundary from Coder</td>
</tr>
</tbody></table>

<p>Put together, the stack above + Coder Agents keeps the control plane, agent orchestration, and source code inside the customer&#39;s infrastructure, and when models are served on OpenShift AI, inference stays there too.</p>

<h2><strong>What you get running Coder on OpenShift</strong></h2>

<p>Organizations in regulated industries require that source code, model usage, and agent activity remain within controlled environments.</p>

<p>AI agents introduce new risks: network access, unmanaged credentials, and lack of auditability<strong>.</strong> OpenShift secures the infrastructure. Coder governs how AI operates within it. Running Coder on OpenShift addresses these risks:</p>

<ul>
<li>AI Gateway centralizes model access and logging</li>
<li>Agent Firewall enforces network boundaries for agents</li>
<li>Coder Agents keeps agent orchestration and credentials in the control plane, with workspaces used only for code execution</li>
</ul>

<p>The result is a system where AI agents operate on enterprise infrastructure with user-level attribution, controlled network access, and complete audit trails.</p>

<h2><strong>Getting started</strong></h2>

<p>Coder runs on OpenShift under the default <code>restricted</code> SCC. No custom SCCs, no privileged containers, and Agent Firewall&#39;s Landjail mode operates without elevated permissions, so it works out-of-the-box on hardened clusters. Deployment is a standard Helm install with a few OpenShift-specific values (Project-assigned UID/GID ranges, <code>Route</code> configuration for workspace access); the full walkthrough lives in the <a href="https://coder.com/docs/install/openshift">Coder on OpenShift install guide</a> and stays current there as OpenShift versions evolve. </p>

<p>To try Coder Agents during Early Access, enable the <code>agents</code> experiment flag, assign the <code>Coder Agents User</code> role, and then configure any supported LLM provider in the Admin settings. See the <a href="https://coder.com/docs/ai-coder/agents/getting-started">Coder Agents documentation</a> for specifics.</p>

<p>Together, Coder and OpenShift provides the infrastructure and governance layer needed for AI to move from test to production.  The <a href="https://coder.com/docs/ai-coder">Coder AI Governance i</a>nstall guide to get a deployment running, and the <a href="https://coder.com/docs/ai-coder">AI Governance documentation</a> covers the full architecture for teams ready to go deeper. For more details <a href="https://coder.com/demo">request a demo</a>.</p>


AI coding agents need governance on hardened infrastructure. Coder on Red Hat OpenShift delivers centralized model access, network controls, and audit trails

Coder on Red Hat OpenShift: AI Agents and Governance for Regulated Environments

<p><em>April 2026 update: This post references Coder Tasks, an earlier approach to orchestrating agent workflows on Coder infrastructure. For the latest capabilities, check out <a href="https://coder.com/docs/ai-coder/agents">Coder Agents</a>.</em></p>

<hr>

<p>Developers live inside their IDE. It is where they read code, write code, debug, review diffs, and think. Every time they leave that environment to open a browser tab, trigger a background workflow, check logs, or monitor an agent, they break flow. Even small context switches add up. And as autonomous AI agents become a bigger part of the development process, that problem compounds. Agents need to be launched, monitored, and managed, and right now, that means leaving the IDE entirely.  </p>

<p>That&#39;s the gap Coder Tasks was built to close. Last year, we introduced <a href="https://coder.com/docs/ai-coder/tasks">Coder Tasks</a> as an interface for running and managing background agentic workflows. Every Task runs in a Coder workspace that has the correct repositories, credentials, and network access. It has been widely used by enterprises as they retain control; all activity is auditable while the developers can use the Coder UI or API to create a Task and monitor its progress. </p>

<p>Until now, Tasks were not available where engineers work most - inside their IDE. That is why we are bringing Coder Tasks directly to the IDE starting with VS Code and its forks like Cursor. Instead of opening a browser to create a Task, monitor progress, or download logs, developers can launch and manage Coder Tasks directly from the IDE.</p>

<p>This keeps engineers in flow. It eliminates context switching to a browser. Developers can stay in their editor and let autonomous workflows run in the background..</p>

<p>Once the Coder extension is installed from the <a href="https://marketplace.visualstudio.com/items?itemName=coder.coder-remote">VS Code marketplace</a>, two Coder icons appear in the sidebar. One icon connects to standard Coder workspaces. The other opens the Coder Tasks view, as shown: </p>

<p><img src="https://www.datocms-assets.com/19109/1772578999-tasks-in-ide2.png" alt=""></p>

<p>From the Tasks view developers can:</p>

<ol>
<li>Select a task template<br></li>
<li>Configure preset parameters<br></li>
<li>Prompt the agent<br></li>
<li>Create and launch the Task</li>
</ol>

<p>After a Task is created, the Task History panel provides control and insight. Developers can:</p>

<ul>
<li>View chat logs from the autonomous workflow<br></li>
<li>Pause or resume the Task workspace<br></li>
<li>Download logs for audit or inspection<br></li>
<li>Delete the Task workspace when the work is complete</li>
</ul>


Introducing Coder Tasks in VS Code

<h2>How it works: A common use case</h2>

<p>Lets say you are in Cursor implementing a feature, and want to have an agent write unit tests for your new API while you start on the frontend. Instead of breaking the flow or switching tabs to an AI chat interface:</p>

<ol>
<li>Initiate: From within the IDE, you go to the Coder Tasks interface.<br></li>
<li>Prompt: You enter a prompt such as “Create a new branch, and write unit tests that cover the basic test scenarios for the new API I implemented” and create the task.<br></li>
<li>Continue: You immediately return to coding your frontend while the task runs in the background.<br></li>
<li>Monitor: At any time, you can check the status (idle/active/errored) or view the last message sent by the agent directly in the extension.<br></li>
<li>Finalize: Once the agent has finished working, you can create a PR from the branch and merge it.</li>
</ol>

<p>The entire experience stays inside your IDE.</p>

<h2>It solves for more than a UI inconvenience</h2>

<p>Many IDE-agent experiences rely on SaaS/remote compute that is disconnected from enterprise command and controls. These agents may not have access to internal services or VPC resources. Or they may require broad external access that admins can’t audit and approve. In either case, productivity suffers or risk increases. </p>

<p>Coder Tasks solve this directly. Every Task runs inside a secure Coder workspace. That workspace is defined as code and hosted in  your own environment inheriting network access, IAM policies, and controls of the hosting environment defined by you. It has access to correct repos, credentials, and internal connectivity already configured.</p>

<p>The IDE is where development happens. As autonomous agents become a bigger part of that work, enterprises need them to run where developers already are, not in a separate tab, not on disconnected SaaS, but inside governed workspaces with the right access and full auditability.</p>

<p>Coder Tasks in VS Code brings those two things together. Developers stay in flow. Admins stay in control.</p>

<p>Install the Coder extension from the VS Code Marketplace and try it today. If you want to see how Coder Tasks fits into your enterprise AI stack, book a <a href="https://coder.com/demo">demo</a>.</p>

<h2>Resources</h2>

<ul>
<li><a href="https://coder.com/docs/ai-coder/tasks">Coder Tasks Documentation</a><br></li>
<li><a href="https://marketplace.visualstudio.com/items?itemName=coder.coder-remote">VS Code Marketplace</a><br></li>
<li><a href="https://coder.com/blog/inside-the-stack-secure-and-scale-ai-coding-agents-with-coder">Coder’s AI Stack</a><br></li>
<li><a href="https://coder.com/blog/every-cursor-needs-a-coder">Unblocking AI Coding Tools in the Enterprise</a></li>
</ul>


Tab switching is a productivity killer. The Coder Tasks plug-in for VS Code-based IDEs including Cursor allows you to trigger Coder Tasks directly from within your IDE.

Don’t Break Flow or Security: Spin Up Coder Tasks Within VS Code Extensions

Don’t Break Flow or Security: Spin Up Coder Tasks Within VS Code

<p><em>April 2026 update: AI Bridge is now AI Gateway, and Agent Boundaries is now Agent Firewall, with no functional changes. This post references Coder Tasks, an earlier approach to orchestrating agent workflows on Coder infrastructure. For current capabilities, see <a href="https://coder.com/docs/ai-coder/agents">Coder Agents</a>.</em></p>

<hr>

<p>Platform teams across industries are exploring how to expand their adoption of AI coding agents, with or without a human in the loop. But we often see the momentum slows down when governance, security, and compliance become part of the equation. We’ve learned from our customers that the difference between a stalled pilot and production-ready AI development comes down to infrastructure. </p>

<p>Coder helps teams address these challenges and unblock AI tools in their workflows. Our customers deploy Coder for a consistent and well-governed enterprise development platform that supports both human and AI agent workflows. </p>

<p>The following technology stack offers a blueprint for deploying AI coding agents securely and effectively in enterprise environments for various industries and use cases. This is how some of the world’s leading teams are running AI agents in production today.</p>


<h2>Essential concepts</h2>

<p>Before we walk through the Coder AI stack, here are the core components that make it work:</p>

<ul>
<li><a href="https://coder.com/docs/v2/latest/templates">Templates</a> are written in Terraform and define the underlying infrastructure that all Coder workspaces run on. Admins create these templates to ensure consistent, reproducible environments for developers and AI agents. Templates may consist of one or multiple files that represent the infrastructure hosting Coder workspaces. It all starts with the templates!</li>
<li>A <a href="https://coder.com/docs/v2/latest/workspaces">Workspace</a> is the development environment where developers or AI agents work inside of. They are provisioned using Terraform, inside of the customer infrastructure. Each Workspace will have one associated user.<br></li>
</ul>


Coder starter templates

<ul>
<li><a href="https://coder.com/docs/ai-coder/tasks">Coder Tasks</a> is an interface that allows you to run and manage coding agents like Claude Code and Amazon Q within a given Workspace. A developer can jump in and steer the agent as required.<br></li>
<li><a href="https://coder.com/docs/ai-coder/security#set-up-agent-boundaries-premium">Agent Boundaries</a>  enable admins to encapsulate agents like Claude Code to run within a separate firewall and add process isolation, preventing agents from accessing resources they are not allowed to.<br></li>
<li><a href="https://coder.com/docs/ai-coder/ai-bridge">AI Bridge</a> acts as a gateway between your users&#39; coding agents or IDEs and LLM providers. By intercepting all AI traffic between these clients and the upstream APIs, AI Bridge can record user prompts, token usage, and tool invocations.<br></li>
</ul>

<p>These building blocks are designed to work as a unified system where infrastructure-as-code definitions flow into governed execution environments and AI interactions pass through centralized security controls. The result is an architecture where platform teams maintain control while developers and AI agents maintain velocity.</p>


<h2>Coder&#39;s blueprint for running AI agents</h2>

<p>To developers and AI agents, all that matters is that they have consistent environments that just work with all the tools they need. No setup friction. No security exceptions. No wondering if the agent can access what it needs or if it&#39;s violating policy.</p>

<p>Coder is cloud, IDE, and agent agnostic which means admins can pick and choose tools that their users like and create standardized templates for faster onboarding. The key components for Coder’s AI-native development platform are illustrated in the diagram below:  </p>


<h2>The AI stack that makes it possible</h2>

<p>Six layers work together to deliver security, consistency, and scale. We&#39;ll walk through them bottom-up, starting with the foundation and building toward where work actually happens.</p>

<h4><em>Infrastructure layer: Deployment flexibility</em></h4>

<p>This is the foundation where the hardware lives, representing any public cloud (AWS, GCP, Azure) or private infrastructure (with VMware, etc). Coder is self-hosted, which means, you can deploy it on any hyperscaler, on-prem, or even in air-gapped datacenters to adhere with potential regulatory requirements. All deployment models can be provisioned with availability, reliability and scale in mind. </p>

<p>Since Coder is configured inside your own infrastructure, it inherits all the access and security policies you define. On top of the infrastructure, you can choose to run  Kubernetes, VMs on EC2, OpenStack, VMware and so on. This abstracts and manages the compute layer Coder runs on so teams don’t need to manage raw infrastructure and can standardize how development environments are provisioned. Most commonly, we see teams deploying Kubernetes clusters for cloud-native workloads and AWS EC2 for monorepo-heavy development environments. </p>

<p>The next three layers are part of Coder software: an AI control plane, a development environment, and an LLM gateway.</p>

<h4><em>AI Control Plane: The Orchestration layer</em></h4>

<p>Coder’s AI control plane serves as the central coordinator to manage Workspace lifecycles, user authentication, agent permissions, Workspace configuration, and external integrations. Both UI &amp; API are supported for template creation and hosting, provisioning, Role-Based Access Control (RBAC), governance, and lifecycle management. </p>

<p>Coder separates who defines environments from who uses them. Admins create and manage Templates while developers use those Templates to launch Workspaces. Templates are responsible for creating and running AI agents within Workspaces. This separation ensures consistency, repeatability, and governance at scale.</p>

<h4><em>Development Environment: Where governance meets execution</em></h4>

<p>This is the most critical part of the stack. Here, Workspace infrastructure spans the resources that AI tools and developers use to read, write, and execute code with pre-configured tools, boundaries, policies, and governance during runtime.  Every Workspace is configured via infrastructure as code, so you can track changes, replicate setups, and keep everything consistent across your fleet. </p>

<p>Agent Boundaries provide network-level enforcement to block domains, subnets, or HTTP verbs and prevent exfiltration inside of your Coder Workspace, ensuring that an AI agent does not operate with the same level of privileges as its human counterpart.</p>

<p>Admins can select any AI tools to run on Coder, whether it’s AI coding assistants like Cursor or Copilot that help developers write, refactor, and document code in real time, or AI coding agents like Claude Code or Codex that autonomously complete scoped tasks such as writing, testing, and committing code changes without step-by-step human guidance.</p>

<p>You can run AI coding agents alone, with users, or just the code assistants depending on the task in point. For example, a team triaging a complex bug may assign initial investigation to Claude Code within a Coder workspace, which analyzes logs, traces the issue, and proposes potential fixes. The developer then jumps into their preferred IDE within the same workspace to review the analysis and complete the implementation. </p>

<p>Both developer and agent have access to source code and tools, but only the developer can access integrations, user credentials, and network access. Agents can operate on code and tooling but can&#39;t impersonate users, access credentials, or reach external networks unless explicitly allowed.</p>

<p>Common Coder Workspaces configurations include:</p>

<ul>
<li><p><strong>Coding Assistants</strong>: These are the most widely adopted form of AI today and often serve as a gateway to more advanced AI coding agents. Coding assistants are great for in-line suggestions, autocompletions, and code explanations and can run inside Coder Workspaces. Coder integrates with agentic IDEs such as <a href="https://registry.coder.com/modules/coder/cursor">Cursor</a>, <a href="https://registry.coder.com/modules/coder/windsurf">Windsurf</a>, and <a href="https://registry.coder.com/modules/coder/zed">Zed</a> to work alongside developers. Template admins can pre-install extensions for coding assistants such as GitHub Copilot and Roo Code. Running these assistants inside Coder workspaces rather than locally provides faster onboarding with out-of-the-box setup, standardized environments, and security through isolated environments. Common use cases include small refactors, generating unit tests, writing inline documentation, and code search and navigation.</p></li>
<li><p><strong>Human + Agent with Coder Tasks</strong>: Best for collaborative development, where a human and an AI agent (e.g., Claude Code) share the same Coder Workspace. Coder provides fully-capable Workspaces to switch to an IDE and take issues to completion for cases where humans pair-program with agents. Developers delegate research and initial implementation to AI, then take over in their preferred IDE to complete the work. The difference is the access level, as in, both developer and their AI agent will have access to source code and tools while developer only would be able to access integrations, user credentials and network access. This ensures that agents can operate on code and tooling, but can’t impersonate users, access credentials, or reach external networks unless explicitly allowed. Common use cases include bug triage and analysis, exploring technical approaches, understanding legacy code, and creating starter implementations.</p></li>
<li><p><strong>Agent-only Task</strong>: Great for AI agents that run independently without user interaction for extended periods of time. Coder Tasks provides isolated, context-based Workspaces with source code, dev tools and services allowing for asynchronous tasks execution. The workspace shell is not exposed externally and is used for managing agent execution such as job orchestration or monitoring. Tasks become available when the template for a workspace has the coder_ai_task resource and coder_parameter named AI Prompt defined in its source code. In other words, any existing template can become a Task template by adding in that resource and parameter. Common use cases include automated code reviews, scheduled data processing, continuous integration tasks, and monitoring and alerting.</p></li>
</ul>

<h4><em>LLM Gateway: Secure model access</em></h4>

<p>Coder AI Bridge is an LLM gateway that interfaces model calls and manages centralized access to MCP servers (e.g., Jira, GitHub) for AI agents and assistants. AI Bridge sits between developers and LLM providers, monitoring AI traffic between these clients and the upstream APIs to record user prompts, token usage, and tool invocations.</p>

<p>AI Bridge centralizes logging of prompts and tool calls, leverages existing identity/RBAC, and supports both centrally configured and local tools. This enables teams to deliver AI development tools responsibly and at scale while ensuring compliance, cost controls, and security policies are enforced. AI Bridge sits between developers and LLM providers, enabling rapid, enterprise-scale adoption of AI tools with built-in governance.</p>

<h4><em>Model Provider</em></h4>

<p>A model provider standardizes access to different AI models, however you can also self-host your own. Common model providers would be the likes of AWS Bedrock, Anthropic, and Open AI. Model providers integrate with AI Bridge providing authentication, observability, and cost management regardless of which models your teams choose.</p>

<h4><em>LLM Models</em></h4>

<p>Coder is agnostic to which LLM model you choose, whether it is a common model provided by major vendors or your own custom models. We support different LLM models through AI Bridge, which allows teams to configure which providers and models their AI agents and assistants use. This setup enables organizations to use LLMs within their own infrastructure while maintaining consistent governance, compliance, and security controls. Research which models you are comfortable with and configure your Coder templates to use them.</p>


<h2>From architecture to action</h2>

<p>Running AI agents in enterprise development environments introduces new risks such as access to sensitive code, potential data exfiltration, policy violations, or unintended changes. The above technology stack will help you address these risks by enforcing clear execution boundaries, scoping permissions per task, logging every interaction for traceability, and supporting hybrid human/agent collaboration with role-based access. </p>

<p>The stack you&#39;ve just seen is how enterprises in financial services, defense, and high tech are running AI agents in production today. The difference between their success and the 40% of AI projects Gartner predicts will be scrapped comes down to infrastructure built for governance from day one.</p>

<p>Here&#39;s what changes when you adopt AI agents with Coder:</p>

<ul>
<li><strong>Security teams</strong> get deterministic controls rather than policy documents hoping developers comply</li>
<li><strong>Platform engineers</strong> get consistent environments rather than snowflake setups that break in production</li>
<li><strong>Developers</strong> get AI acceleration rather than AI friction and endless security exceptions</li>
<li><strong>Leadership</strong> gets production deployment rather than stalled pilots and wasted investment</li>
</ul>

<p>Ready to see how this AI stack works in your environment? <a href="https://coder.com/trial">Try out Coder</a> and explore how Coder can strengthen your development workflows by unlocking AI tools in your enterprise.</p>

<h2>Resources</h2>

<p><a href="https://coder.com/docs">Coder Documentation</a><br>
<a href="https://coder.com/docs/user-guides/workspace-access">Access Coder Workspaces</a><br>
<a href="https://coder.com/docs/ai-coder/tasks">Coder Tasks</a><br>
<a href="https://coder.com/docs/ai-coder/security">Security and Boundaries</a><br>
<a href="https://registry.coder.com/modules">Coder Registry</a><br>
<a href="https://coder.com/pricing">Coder Premium vs Community packaging</a></p>


Platform teams across industries are exploring how to expand their adoption of AI coding agents, with or without a human in the loop. Coder helps teams address common AI challenges and unblock AI tools to improve productivity and experiences for users.

blog image

Inside the Stack: Secure and Scale AI Coding Agents with Coder

<p>Kiro IDE by Amazon joins the growing lineup of agentic IDEs built on VS Code. Kiro takes a different approach, emphasizing spec-driven development — planning before coding. It’s currently powered by Claude Sonnet, with more models on the roadmap. </p>

<p>It automatically generates detailed requirements, design documentation, and discrete tasks upfront, giving you clear visibility into what the AI intends to build before writing any code. Other great features include agent hooks, fine-grained steering, and multi-modal inputs, making Kiro a compelling choice for teams looking to build with structure and control.</p>

<p>Using Kiro? Ask your Coder admin to add the <a href="https://registry.coder.com/modules/coder/kiro">Kiro module</a> to your template. This enables an “Open in Kiro” button inside your workspace. </p>

<p>Use the <a href="https://github.com/coder/vscode-coder">Coder Remote VS Code Extension</a> and <a href="https://open-vsx.org/extension/jeanp413/open-remote-ssh">open-remote-ssh extension</a> for establishing connections to Coder workspaces.</p>


<p>Why use Kiro with Coder? Integrating Kiro with Coder workspaces unlocks key benefits for admins, including:</p>

<h3>Remote vs local setup</h3>

<p>Launch the Kiro client locally and connect to Coder&#39;s cloud dev environments. Developers skip complex local setup — getting onboarded fast, with collaboration built in. Coder is fully self-hosted which means it inherits all security policies of your own environment further helping with any air-gapped or industry accreditation requirements. </p>

<h3>Standardization and governance</h3>

<p>Coder employs Terraform to create templates for reproducible, development environments called Coder workspaces, helping with version control and standardization of both your infrastructure and the environment. Admins can add modules like Kiro from the <a href="https://registry.coder.com/modules">Coder Registry</a>, with audit logging, identity federation, and tool usage controls built in.</p>

<h3>Fast and consistent MCP tooling</h3>

<p>With Coder’s Kiro integration, configure MCP servers directly into template code, adding organization-specific context — no login delays, no workflow disruptions. It scales with your team, Kiro automatically detects and loads your MCP servers from the configuration file. This approach speeds up developer onboarding without needing any manual configuration steps every time.</p>


<h2>Configure MCP server for Kiro</h2>

<p>Provide a JSON-encoded string via the mcp input. When set, the module writes the value to ~/.kiro/settings/mcp.json using a coder_script on workspace start. </p>

<p>After you apply the configuration to your environment, every workspace is automatically set up to let Kiro connect to MCP servers. This requires no developer intervention and ensures consistency across all environments.</p>

<p>The following example configures Kiro to use the GitHub MCP server with authentication facilitated by the <a href="https://coder.com/docs/admin/external-auth#configure-a-github-oauth-app">coder<em>external</em>auth</a> resource.</p>


<p>Kiro supports both “vibe coding” and structured, spec-driven workflows. Coder makes it easy to deploy Kiro and other tools in a secure, enterprise-ready environment that scales with your team.</p>


<h2>Getting started with Kiro IDE in Coder</h2>

<ul>
<li>Add Kiro IDE in under a minute: <a href="https://registry.coder.com/modules/coder/kiro">https://registry.coder.com/modules/coder/kiro</a> </li>
<li>MCP configuration and best practices for Kiro: <a href="https://kiro.dev/docs/mcp/configuration">https://kiro.dev/docs/mcp/configuration</a> </li>
<li>For any inquiries or support, please <a href="https://github.com/coder/registry">open an issue here</a> or <a href="https://discord.gg/coder">join our Discord community</a>.</li>
</ul>


With Coder's Kiro integration, configure MCP servers directly into template code, adding organization-specific context — no login delays, no workflow disruptions.

AI-native Development

Secure Environments for 2,500+ Developers

Suman Bisht

Explore the latest from Suman

Coder on Red Hat OpenShift: AI Agents and Governance for Regulated Environments

Don’t Break Flow or Security: Spin Up Coder Tasks Within VS Code

Inside the Stack: Secure and Scale AI Coding Agents with Coder

Coder Brings Governance to Kiro IDE with Centralized MCP

Related content

Using Claude Code with Coder Workspaces

Introducing Coder Agents

Self-Hosted Doesn’t Always Mean What It Implies