Canvas + OpenClaw

How to connect Canvas to OpenClaw

Connect Canvas to OpenClaw through Agent200's MCP server: managed OAuth, ready-made tools and one API key.

On this page
  1. Overview
  2. Before you start
  3. Setup steps
  4. Test your connection
  5. Instructions for AI agents
  6. Canvas tools
  7. How it works
  8. Troubleshooting

Overview

This guide connects Canvas to OpenClaw through Agent200's MCP server. When you finish, OpenClaw can search, read and act in Canvas with the tools Agent200 provides, whenever a request needs them.

You add one MCP server to OpenClaw with your Agent200 API key. Your Canvas account is connected through managed OAuth: you approve access once on Canvas's own page, and Agent200 keeps the connection working.

Connector
Canvas
Works with
OpenClaw ยท Agent app, through MCP
Authentication
Managed OAuth, per end user
Tools
7 tools
MCP server
https://api.agent200.com/mcp/

Before you start

  • An Agent200 API key. Create one on the API keys page of the dashboard.
  • Canvas enabled for your project. In the dashboard's Connectors settings, with the endpoints and scopes your agent needs.
  • An end user with a Canvas account. They approve access once, on Canvas's own page.
  • OpenClaw, with custom MCP servers allowed. OpenClaw is open source and self-hosted: anyone running it can add MCP servers, from the CLI, the config file or the Control UI.

Connect Canvas to OpenClaw

  1. 1

    Step 1: Create your Agent200 API key

    Open the API keys page in the Agent200 dashboard and choose Create an API key. You add it to OpenClaw in the next step, as "Authorization: Bearer <your key>". One key covers Canvas and every other connector enabled for your project. Keep it out of chats and source control.

  2. 2

    Step 2: Store your key in ~/.openclaw/.env

    OpenClaw fills ${AGENT200_API_KEY} in its config from ~/.openclaw/.env, so the key never sits in the config itself.

    ~/.openclaw/.env
    MCP: OpenClaw key in ~/.openclaw/.env
    AGENT200_API_KEY=your-agent200-api-key
  3. 3

    Step 3: Save the server

    openclaw mcp set takes one JSON object. Always set the transport to streamable-http: without it OpenClaw uses SSE. Keep the single quotes so your shell leaves ${AGENT200_API_KEY} for OpenClaw. In the Control UI, add it under Settings > MCP, then put the header in the config editor on the same page.

    Terminal
    MCP: OpenClaw save the server from the CLI
    openclaw mcp set agent200 '{"url":"https://api.agent200.com/mcp/","transport":"streamable-http","headers":{"Authorization":"Bearer ${AGENT200_API_KEY}"}}'
  4. 4

    Step 4: Probe the server

    Open a live connection and list the tools Agent200 offers. If a Gateway runs elsewhere, reload or restart it.

    Terminal
    MCP: OpenClaw probe the server
    openclaw mcp doctor agent200 --probe
    openclaw mcp status --verbose
  5. 5

    Step 5: Connect your Canvas account

    The first time you ask OpenClaw for something in Canvas, Agent200 returns an authorization link. Open it and approve access on Canvas's own page. Agent200 stores the connection and keeps it working; your password never reaches OpenClaw or Agent200.

Test your connection

Run one request to check everything is wired up.

Ask

What assignments are due this week in Biology 101

What should come back

  • A result from that user's own Canvas account.
  • Behind the scenes, OpenClaw calls canvas_list_courses.

If you get an authorization link instead, the user has not connected Canvas yet. Show them the link, then run the request again.

Get your API key

Create a key in the Agent200 dashboard. It reaches Canvas and every other connector enabled for your project.

Read the SDK docs

Instructions for AI agents

Using a coding agent? Paste this brief into it, or give it the address of this page. It lists what to do, in order, to connect Canvas to OpenClaw.

Agent brief
Goal: connect Canvas to OpenClaw with Agent200, through its MCP server.
Guide: https://www.agent200.com/connectors/canvas/framework/openclaw/

1. The user needs an Agent200 API key. If they have none, send them to https://agent200.com/platform/apikeys/ to choose "Create an API key". They paste it into OpenClaw's settings themselves: never ask for it in chat and never write it into files you share.
2. Make sure Canvas is enabled for their project in the Agent200 dashboard (Connectors), with the tools the task needs.
3. Add Agent200's MCP server to OpenClaw: URL https://api.agent200.com/mcp/, transport streamable HTTP, header "Authorization: Bearer <the user's Agent200 API key>". Follow the setup steps on the guide page for OpenClaw's exact menus or commands.
4. The first Canvas request returns an authorization link if the user has not connected Canvas yet: show it to them and continue once they have approved access on Canvas's own page.
5. Check it works with this request: "What assignments are due this week in Biology 101"
   What should come back:
   - A result from that user's own Canvas account.
   - Behind the scenes, OpenClaw calls `canvas_list_courses`.
6. If a tool is missing or a call is refused, check the project's Connectors settings in the dashboard and the access the user granted in Canvas, then retry.

The whole guide is also available as Markdown, for agents and LLMs: /connectors/canvas/framework/openclaw.md

Canvas tools OpenClaw can call

  • List courses

    Read
    canvas_list_courses

    List the user's courses.

  • List assignments

    Read
    canvas_list_assignments

    List the assignments in a course.

  • List modules

    Read
    canvas_list_modules

    List the modules in a course.

  • List files

    Read
    canvas_list_files

    List the files in a course.

  • List submissions

    Read
    canvas_list_submissions

    List the submissions for an assignment.

  • Post an announcement

    Write
    canvas_create_announcement

    Post an announcement to a course.

  • Grade a submission

    Write
    canvas_grade_submission

    Set a grade or comment on a submission.

How it works

  1. 01

    You ask OpenClaw

    In your own words, for something in Canvas.

  2. 02

    OpenClaw picks a tool

    It sees the Canvas tools on the Agent200 MCP server and calls the one the request needs.

  3. 03

    Agent200 runs the call

    With your own Canvas connection, within the tools and scopes your project allows.

  4. 04

    OpenClaw answers

    It gets the result back from the MCP server and uses it in its reply or its next step.

Each user's own connection. Connections belong to individual end users. A request for one user never reaches another user's Canvas account.

Permissions at every level. Your project settings decide what the agent may request, the end user decides what they authorize, and Canvas enforces what its API grants.

Troubleshooting

The connection fails right after adding the server

Check the transport is streamable-http: when it is missing OpenClaw uses SSE, which Agent200 does not serve.

Authentication refused (401)

Check headers.Authorization is Bearer ${AGENT200_API_KEY} and that AGENT200_API_KEY is set for the process running OpenClaw.

The doctor warns about a literal header value

Replace the pasted key with ${AGENT200_API_KEY} and move the key to ~/.openclaw/.env.

Changes do not reach your agent

A Gateway or agent running elsewhere needs its own reload or a restart.

OpenClaw shows a Canvas authorization link

You have not connected Canvas yet, or the connection needs to be approved again. Open the link, approve access on Canvas's own page, then ask OpenClaw again.

A tool call is refused or returns less than expected

Agent200 runs only what your project allows. Check that Canvas and the tools you need are enabled in the dashboard's Connectors settings, and that you granted Canvas the access the task needs.

The API key was pasted into a chat or committed

Create a new key on the API keys page, replace it in OpenClaw's MCP settings, and stop using the old one. Paste the key only into OpenClaw's settings, never into a chat.

Use Canvas with another framework

200 OK

Canvas, connected to OpenClaw.
One SDK. One API key.

Agent200 handles the OAuth, the tokens and every tool call. You keep your agent, your model and your code.

Book a demo.
See it in action.

Tell us who you are, then pick a time for a 1:1 session with an expert from our team.

We use your details only to email you about Agent200. See the Privacy Policy.

Pick a time. We'll take it from there.

A 1:1 session with an expert from our team, about what you are building.

Open in Calendly (opens in a new tab)