opensource/coder
3
0
Fork 0
mirror of https://github.com/coder/coder.git synced 2026-09-21 12:44:32 +08:00
Code Issues Packages Projects Releases Wiki Activity

docs: add Coder Agents architecture deep-dive page (#22625)

Browse Source 
Adds a new child page under **Coder Agents**
(`/docs/ai-coder/agents-architecture`) that explains how the agent in
the control plane communicates with workspaces.

## Core message

The Coder Agent interacts with workspaces using the exact same
connection path as a developer's IDE, web terminal, or SSH session — no
special protocol, no sidecar, no new ports.
This commit is contained in:
Matt Vollmer
2026-03-04 19:17:02 -05:00
committed by GitHub
parent 30d534b36b
commit 28d99e8afb
5 changed files with 278 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/ai-coder/agents-architecture.md
+255
View File
@@ -0,0 +1,255 @@
# Architecture
Coder's AI agent interacts with workspaces over the same
connection path as a developer's IDE, web terminal, and SSH session already
use. There is no sidecar process and no new network paths. If your developers
can already connect to their workspaces, the agent can too.
## Architecture at a glance
Three components are involved in every agent interaction:
1. **The control plane** runs the agent loop. It receives prompts, streams them
to the LLM provider, interprets tool calls, and dispatches them to
workspaces.
1. **The LLM provider** (Anthropic, OpenAI, Google, Azure, AWS Bedrock, or any
OpenAI-compatible endpoint) performs model inference. It never communicates
with the workspace directly.
1. **The workspace** is standard compute infrastructure. It runs shell commands,
reads and writes files, and executes processes — exactly what occurs when a
developer connects via their IDE.
![Architecture diagram ](../images/guides/ai-agents/agent-loop-detailed.png)
## The same connection your IDE uses
This is the key architectural insight: the agent reaches into a workspace
over the same Tailnet tunnel that a developer's tools already use.
When a developer opens a web terminal in the Coder dashboard, connects via
VS Code Remote, or runs `coder ssh`, the traffic follows this path:
1. The client connects to the control plane.
1. The control plane routes the connection through its internal Tailnet node.
1. The connection reaches the workspace daemon over a DERP relay or
direct peer-to-peer link.
1. The workspace daemon handles the request — spawning a shell,
forwarding a port, or serving a file.
When the agent executes a tool call — reading a file, running a command,
writing code — it follows the same tunnel:
1. The agent loop in the control plane issues a tool call.
1. The control plane routes the call through its internal Tailnet node.
1. The call reaches the workspace daemon over the same DERP relay or
peer-to-peer link.
1. The workspace daemon handles the request via its HTTP API — reading a file,
starting a process, or writing content.
The underlying tunnel is identical. IDE connections use SSH, web terminals use
a WebSocket protocol, and the agent uses the workspace daemon's HTTP API — but
all three traverse the same Tailnet connection and rely on the same security
boundary. No additional ports or network paths are introduced.
### No inbound ports
The workspace daemon always dials _out_ to the control plane — never
the reverse. The control plane then uses that established tunnel to reach back
in. This means:
- The workspace needs no inbound ports or exposed services.
- You can block all inbound traffic to the workspace.
- The only required outbound connection from the workspace is to the control
plane itself.
This is unchanged from how workspaces already operate in Coder. Enabling
Coder Agents does not change your workspace network requirements.
## The agent loop
When a user submits a prompt, the control plane processes it as a background
job:
1. The prompt is saved to the database and the chat is marked `pending`.
1. The control plane picks up the chat and marks it `running`.
1. The control plane streams the conversation to the configured LLM provider.
1. The model responds with text, reasoning, or tool calls.
1. If the response includes tool calls, the control plane executes them
(connecting to the workspace as needed) and returns the results to the model.
1. Steps 3–5 repeat until the model produces a final response with no further
tool calls.
1. The chat is marked `waiting` for the next user message.
This loop runs inside the control plane process. There is no separate service
to deploy — it is part of the same binary that serves the dashboard and API.
### Context compaction
As conversations grow, the agent automatically summarizes older context to stay
within the model's context window. When token usage exceeds a threshold, the
agent generates a compressed summary and inserts it as a new message. Earlier
messages remain in the database and are still visible to users, but are excluded
from the model's context window. This happens transparently and keeps
long-running sessions productive.
### Message queuing
Users can send follow-up messages while the agent is actively working. Messages
are queued in the database and delivered when the agent completes its current
turn — the full sequence of steps until the model stops calling tools. There is
no need to wait for a response before providing additional context or
redirecting the agent.
## Tool execution
Tools are how the agent takes action. Each tool call from the LLM translates to
a concrete operation — either inside a workspace or within the control plane
itself.
### Workspace connection lifecycle
The connection to a workspace is **lazy**. It is not established when a chat
starts — only when something needs to reach the workspace. This is typically
triggered by the first tool call that requires workspace access. Once
established, the connection is cached and reused for the duration of that chat
session.
Chats that don't need workspace access (answering questions, planning an
approach, discussing architecture) never provision or connect to a workspace.
### Workspace tools
These tools execute inside the workspace via the workspace daemon's HTTP API.
They traverse the same Tailnet tunnel used by web terminals and IDE connections.
| Tool | What it does |
|------------------|--------------------------------------------------------------------|
| `read_file` | Reads file contents with line-number pagination. |
| `write_file` | Writes content to a file. |
| `edit_files` | Performs atomic search-and-replace edits across one or more files. |
| `execute` | Runs a shell command (foreground or background). |
| `process_output` | Retrieves output from a background process. |
| `process_list` | Lists all tracked processes in the workspace. |
| `process_signal` | Sends a signal (SIGTERM or SIGKILL) to a background process. |
### Platform tools
These tools run entirely within the control plane. They do not require a
workspace connection.
| Tool | What it does |
|--------------------|-------------------------------------------------------------------|
| `list_templates` | Browses available workspace templates, sorted by popularity. |
| `read_template` | Gets template details and configurable parameters. |
| `create_workspace` | Creates a workspace from a template and waits for it to be ready. |
### Orchestration tools
These tools manage sub-agents — child chats that work on independent tasks in
parallel.
| Tool | What it does |
|-----------------|--------------------------------------------------------------|
| `spawn_agent` | Delegates a task to a sub-agent with its own context window. |
| `wait_agent` | Waits for a sub-agent to finish and collects its result. |
| `message_agent` | Sends a follow-up message to a running sub-agent. |
| `close_agent` | Stops a running sub-agent. |
## What runs where
Understanding the split between the control plane and the workspace is central
to the security model.
| Responsibility | Where it runs | Details |
|---------------------|---------------|---------------------------------------------------------------------------|
| Agent loop | Control plane | Prompt processing, tool dispatch, step iteration. |
| LLM inference | LLM provider | The control plane streams requests to the external provider. |
| Chat state | Control plane | All messages, token usage, and status stored in the database. |
| Git authentication | Control plane | Uses existing Coder external auth (GitHub, GitLab, Bitbucket). |
| User identity | Control plane | Every action is tied to the user who submitted the prompt. |
| Model/prompt config | Control plane | Administrators configure providers, models, and system prompts centrally. |
| File read/write | Workspace | The workspace file system is the source of truth for code. |
| Shell execution | Workspace | Commands run in the workspace's environment with its packages and tools. |
| Git operations | Workspace | Commits, pushes, and branch management happen inside the workspace. |
| Build and test | Workspace | Compilation, test suites, and dev servers run on workspace compute. |
The workspace has **zero AI awareness**. There are no LLM API keys, no agent
processes, and no AI-specific software installed. If you inspect a workspace
created by the agent, it looks identical to one a developer created
manually.
## Chat state and persistence
All chat data is stored in the control plane database, not in the workspace.
- **Chat metadata** — status, owner, associated workspace, timestamps, and
parent/child relationships for sub-agents.
- **Messages** — every message (user, assistant, tool calls, tool results) is
stored as a separate record with role, content, and token usage.
- **Compressed context** — when the agent compacts the conversation, summaries
are stored with a compression flag so the original context budget is
preserved.
- **Queued messages** — follow-up messages sent while the agent is working are
held in a queue and delivered in order.
Because state lives in the database:
- Chat history survives workspace stops, rebuilds, and deletions.
- An administrator can inspect any chat for audit or debugging.
- The agent can resume work by targeting a new workspace and continuing from the
last git branch or checkpoint.
## Security implications
The control plane architecture has direct consequences for how you secure AI
coding workflows.
### No API keys in workspaces
LLM provider credentials exist only in the control plane. The workspace never
sees them. There is nothing for a developer, a compromised dependency, or a
rogue process to exfiltrate.
### Workspaces can be fully network-isolated
Because the workspace does not need to reach any LLM provider, you can restrict
its network access to only:
- The control plane (required for the workspace daemon to function).
- Your git provider (for push/pull operations).
Everything else can be blocked. The AI functionality comes from the control
plane, not from the workspace's network.
> [!TIP]
> For sensitive environments, create dedicated templates for agent workloads
> with stricter egress rules than your standard developer templates. Because
> the AI comes from the control plane, these templates do not need any
> outbound access to LLM providers.
### Centralized enforcement
Administrators control which models are available, the system prompt, and tool
configuration from the control plane. Developers can select from the set of
admin-enabled models when starting or continuing a chat, but cannot add their
own providers or override system prompts or tool permissions. When an
administrator removes a model or modifies the system prompt, the change applies
to all agent sessions immediately.
### User identity on every action
Every action the agent takes — PRs opened, code committed, commands executed —
is tied to the user who submitted the prompt. There is no shared bot account or
anonymous identity. If a developer submits a prompt that results in a pull
request, that pull request is attributed to them via the git authentication
already configured in your Coder deployment.
## Scaling and resource impact
The control plane overhead for Coder Agents is minimal. The heavy computation
happens elsewhere:
- **LLM inference** runs on the external provider's infrastructure.
- **File I/O, builds, and tests** run on workspace compute.
- **The control plane** primarily proxies streaming responses and dispatches
tool calls over existing network connections.
docs/ai-coder/agents.md
+15 -16
View File
@@ -7,8 +7,7 @@
Coder Agents is a chat interface and API for delegating development work and research to coding agents in your Coder deployment. Developers describe the work they want done, and Coder Agents handles selecting a template, provisioning a workspace, and executing the task.
Coder Agents includes its own self-hosted AI coding
agent that runs the agent loop directly within the Coder control plane
(`coder server`).
agent that runs the agent loop directly within the Coder control plane.
It is not a wrapper around third-party agent tools like Claude Code
or Codex.
@@ -50,7 +49,7 @@ other editor to review, refine, and complete work that the agent produces.
## How it works
The agent loop runs inside `coder server`, the Coder control plane. When a user
The agent loop runs inside [the control plane](./agents-architecture.md). When a user
submits a prompt, the control plane:
1. Sends the prompt to the configured LLM provider (Anthropic, OpenAI, Google,
@@ -58,7 +57,7 @@ submits a prompt, the control plane:
1. Receives the model's response, which may include tool calls such as reading
files, writing code, or running shell commands.
1. Executes tool calls by connecting to a Coder workspace over the existing
workspace daemon connection — the same path used for web terminals, port
workspace connection — the same path used for web terminals, port
forwarding, and IDE access.
1. Returns tool results to the model and continues the loop until the task is
complete.
@@ -67,14 +66,14 @@ The workspace itself has no knowledge of AI. It is standard compute
infrastructure — there are no LLM API keys, no agent harnesses, and no special
software installed. All intelligence lives in the control plane.
![Architecture diagram showing the Coder server in the center, with arrows out to LLM providers and arrows to workspaces](../images/guides/ai-agents/agent-loop.png)
![Architecture diagram showing the control plane in the center, with arrows out to LLM providers and arrows to workspaces](../images/guides/ai-agents/agent-loop.png)
<small>The agent loop runs in the Coder server. It makes outbound requests to LLM
<small>The agent loop runs in the control plane. It makes outbound requests to LLM
providers and connects to workspaces only when tool execution is needed.</small>
### Automatic workspace provisioning
Not every chat requires a workspace. The agent runs in `coder server` and can
Not every chat requires a workspace. The agent runs in the control plane and can
answer questions, discuss architecture, or plan an approach without any
infrastructure. Workspaces are only provisioned when the agent needs to take
action — reading code, running commands, or editing files.
@@ -226,15 +225,15 @@ the workspace.
Coder Agents is a new approach that differs from
[Coder Tasks](./tasks.md) in several ways:
| Aspect | Coder Agents | Coder Tasks |
|---------------------|--------------------------------------------|----------------------------------------------------------------|
| Agent execution | Runs in the control plane (`coder server`) | Runs inside the workspace |
| Agent harness | Built-in, no installation needed | Requires Claude Code, Codex, or similar installed in workspace |
| API keys | Stored in control plane only | Injected into workspace environment |
| Chat state | Persisted in database | Stored in workspace |
| Workspace selection | Automatic, based on task description | Manual, user selects template |
| Sub-agents | Built-in parallel delegation | Not supported |
| Modern chat UI | Native chat with diffs, queuing | Terminal-based interface |
| Aspect | Coder Agents | Coder Tasks |
|---------------------|--------------------------------------|----------------------------------------------------------------|
| Agent execution | Runs in the control plane | Runs inside the workspace |
| Agent harness | Built-in, no installation needed | Requires Claude Code, Codex, or similar installed in workspace |
| API keys | Stored in control plane only | Injected into workspace environment |
| Chat state | Persisted in database | Stored in workspace |
| Workspace selection | Automatic, based on task description | Manual, user selects template |
| Sub-agents | Built-in parallel delegation | Not supported |
| Modern chat UI | Native chat with diffs, queuing | Terminal-based interface |
## Product status
docs/images/guides/ai-agents/agent-loop-detailed.png
BIN
View File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 304 KiB

docs/images/guides/ai-agents/agent-loop.png
BIN
View File
Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 83 KiB

After

Width:  |  Height:  |  Size: 82 KiB

docs/manifest.json
+8 -1
View File
@@ -1185,7 +1185,14 @@
{
"title": "Coder Agents",
"description": "Self-hosted agent by Coder",
"path": "./ai-coder/agents.md"
"path": "./ai-coder/agents.md",
"children": [
{
"title": "Architecture",
"description": "How the agent in the control plane communicates with workspaces",
"path": "./ai-coder/agents-architecture.md"
}
]
}
]
},