ClickUp + OpenAI Agents SDK
How to connect ClickUp to OpenAI Agents SDK
Give OpenAI Agents SDK access to ClickUp through Agent200: managed OAuth, ready-made tools and one API key.
On this page
Overview
Build an OpenAI Agents SDK agent that reads an identified ClickUp task through the tools exposed by your Agent200 connection. Limit the agent's tool selection to the read operations needed for the briefing. Keep the MCP client lifecycle and the application's end-user authorization separate from model instructions.
The Agent200 SDK examples on this page are conceptual. Use the actual runtime tool definitions and supported client interfaces supplied to your application. No package installation command, parameter mapping or end-to-end runtime test is implied.
Product reference: OpenAI Agents SDK documentation. For task identity and response scope, see ClickUp’s task reference.
- Connector
- ClickUp
- Works with
- OpenAI Agents SDK · Agent framework
- Authentication
- Managed OAuth, per end user
- You need
- An Agent200 API key
- Tools
- 8 tools
Before you start
- An Agent200 API key. Create one on the API keys page of the dashboard.
- ClickUp enabled for your project. In the dashboard's Connectors settings, with the endpoints and scopes your agent needs.
- An end user with a ClickUp account. They approve access once, on ClickUp's own page.
- A OpenAI Agents SDK agent in your project. Agent200 gives it the tools; your code keeps the orchestration.
Connect ClickUp to OpenAI Agents SDK
-
1
Step 1: Create your Agent200 API key
Create a key on the API keys page. Enable ClickUp and the read operations needed for this task in your project. Keep credentials out of chats and source control.
.envConceptual exampleStore the Agent200 API key as an environment variable # Keep the key out of your code, prompts and chats. export AGENT200_API_KEY="paste-your-key-here" -
2
Step 2: Define the task-reading contract
Select the existing ClickUp task-read operation from the tool definitions supplied by your runtime. Validate the selected task ID and map only documented inputs and outputs. Keep credentials on the server.
-
3
Step 3: Connect the end user's ClickUp account
Have the end user complete the ClickUp authorization flow. Use the same authenticated application user for connection and later requests. Confirm that the user can open the task selected for the first read.
connect.jsConceptual exampleLet an end user connect their account (managed OAuth) // Each end user authorizes their own ClickUp account, once. const connection = await agent200 .user("user_123") .connect("clickup"); if (connection.authorizationUrl) { // Show this link to your user; Agent200 handles the rest. console.log(connection.authorizationUrl); } -
4
Step 4: Build the read-only briefing flow
Use the selected task’s returned details to produce the briefing. Handle missing fields and refused calls explicitly. The following snippet is a conceptual illustration of the site’s Agent200 flow; adapt the implementation to your real runtime contract.
openai-agents-sdk-clickup.jsConceptual exampleGuide: hand the connector's tools to an agent framework // The ClickUp tools for this user, ready for a OpenAI Agents SDK agent. const tools = await agent200 .user("user_123") .tools(["clickup"]); // Give `tools` to your OpenAI Agents SDK agent. // Every call it makes runs through Agent200 with the user's own connection.
Test your connection
Run one request to check everything is wired up.
Ask
Read the ClickUp task selected by this application user, then summarize the returned assignees, due date and status. Ask for a missing task ID and return access failures clearly. Do not change the task.
What should come back
- The selected task is identified clearly.
- The answer preserves its returned status, includes all returned assignees and reports its due date when available.
- Missing data and access failures are stated clearly; no ClickUp records or comments are changed.
Use a task the connected user can open in the authorized Workspace. Check the runtime schema if using a custom task ID. Report missing data and preserve the task’s returned status label.
Get your API key
Create a key in the Agent200 dashboard. It reaches ClickUp and every other connector enabled for your project.
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 ClickUp to OpenAI Agents SDK.
Goal: connect ClickUp to OpenAI Agents SDK with Agent200.
Guide: https://www.agent200.com/connectors/clickup/framework/openai-agents-sdk/
1. Get an Agent200 API key. If the developer has none, send them to https://agent200.com/platform/apikeys/ and ask them to choose "Create an API key". Read the key from the AGENT200_API_KEY environment variable. Never ask for it in chat, and never write it into code, prompts or logs.
2. Make sure ClickUp is enabled for the project in the Agent200 dashboard (Connectors), with the endpoints and scopes the task needs.
3. Add the Agent200 SDK to the project and create a client with that key.
4. Give every request a stable id for the end user. Agent200 uses only that user's own connections.
5. Connect the user's ClickUp account through Agent200. If Agent200 returns an authorization link, show it to the user and continue once they have approved access on ClickUp.
6. Load the user's ClickUp tools ("clickup") from Agent200 and give them to the OpenAI Agents SDK agent. Every call it makes runs through Agent200.
7. Check it works with this request: "Read the ClickUp task selected by this application user, then summarize the returned assignees, due date and status. Ask for a missing task ID and return access failures clearly. Do not change the task."
What should come back:
- The selected task is identified clearly.
- The answer preserves its returned status, includes all returned assignees and reports its due date when available.
- Missing data and access failures are stated clearly; no ClickUp records or comments are changed.
8. 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 ClickUp, then retry.
The code on the guide page is a conceptual example of the SDK. SDK reference: https://docs.igpt.ai/
The whole guide is also available as Markdown, for agents and LLMs: /connectors/clickup/framework/openai-agents-sdk.md
ClickUp tools OpenAI Agents SDK can call
-
Read a task
Readclickup_get_taskRetrieve a task’s description, status, assignees and due date when available.
-
List tasks
Readclickup_list_tasksRead tasks in a selected List, with attention to pagination and the requested task filters.
-
Read task comments
Readclickup_get_task_commentsRead comments on a task to add discussion context to a handoff.
-
Read a List
Readclickup_get_listRead a List’s details to understand the container for the selected work.
-
List Lists in a Folder
Readclickup_list_listsFind the Lists inside a selected Folder.
-
List folderless Lists
Readclickup_list_folderless_listsFind Lists that sit directly in a Space rather than inside a Folder.
-
List Folders
Readclickup_list_foldersRead the Folders in a selected Space to locate the work the user requested.
-
List Spaces
Readclickup_list_spacesRead the Spaces available in a selected Workspace.
No tool matches that search.
How it works
-
01
Your app sends the request
It calls Agent200 with the model, instructions, the input, the end user's id and the ClickUp tools.
-
02
Your OpenAI Agents SDK agent picks a tool
It calls a ClickUp tool when its plan needs one, with the arguments it chooses.
-
03
Agent200 runs the call
It uses that user's own ClickUp connection, within the endpoints and scopes your project allows, and returns the result to the model.
-
04
Your app gets the result
Agent200 returns the completed result. Your app decides when the agent runs and what happens next.
Each user's own connection. Connections belong to individual end users. A request for one user never reaches another user's ClickUp account.
Permissions at every level. Your project settings decide what the agent may request, the end user decides what they authorize, and ClickUp enforces what its API grants.
Troubleshooting
The task cannot be found or read
Check the selected task and authorized ClickUp Workspace. Distinguish native task IDs from custom IDs and use the inputs your tool actually supports. Report a refused call instead of guessing a task or a successful result.
The answer names only one person or changes the status meaning
Compare the answer with the returned record. Keep every returned assignee and the recorded status label. A missing field is not evidence that a task is unassigned or complete.
The List summary seems incomplete
Check pagination and the task filters available in the runtime. State the result scope and avoid treating one page as the full List. See ClickUp’s List task reference.
FAQ
Which SDK is responsible for the agent?
The OpenAI Agents SDK runs your agent. Agent200 supplies the service connection and tools. OpenAI documents Streamable HTTP MCP clients and tool filtering; use its current API reference for the client code. The Agent200 snippets on this site illustrate the architecture and do not define an official package or adapter API.
Can I filter the agent to task reads only?
Select the task-read tools actually advertised by the MCP connection and use the SDK’s supported filtering mechanism. Do not expose unrelated operations merely because the server offers them. Check the returned task before extending the workflow to Lists or comments.
