Claude Code
“Use my product-video app to create a new version for this week’s launch.”
Bring MaxDay into the same conversation where you plan and build your campaign tooling.
MAXDAY / MAXDAY APIYour AI assistant.
Your AI assistant.
Your content engine.
Let Claude Code or Codex run your MaxDay apps. Turn a tested workflow into an API your assistant can call—with new inputs for every product, post, or campaign.
MAXDAY / APP RUNS
A STARTING PROMPT FOR CLAUDE CODE
Use the MaxDay API documentation at
https://maxday.ai/openapi/docs/get-start.
Use MAXDAY_API_KEY from my server environment.
List my projects and published apps. Read the
selected app's inputs before preparing a run.
Run the app with my campaign inputs, then poll
until it reaches a terminal state. Return the
output URLs and report any errors.
Keep my API key out of code, logs, and output.
Do not automatically retry run creation.Claude Code / CodexPrepare your campaign inputs
Your MaxDay appRun your published workflow
Generated assetsGet the output URLs
FROM INSTRUCTION TO OUTPUT
You choose the process. Your assistant handles the run.
Give your coding assistant the API docs and the campaign brief. It can discover your published apps, prepare a request, and retrieve the results.
Codex
“Read the app’s inputs, run it for this campaign, and save the output URLs.”
Let Codex connect the app to your scripts and keep track of the results.
Connect through standard HTTP API calls. Your assistant uses the credentials and app access you provide.
A REAL API FOR REPEATABLE WORK
One app. New inputs. Another run.
Build the workflow visually in MaxDay, then call it from your own code. Every run belongs to a project and returns a run ID you can track.
- 01
Choose a published app
List your projects and apps. Read the app details for its current inputs, outputs, steps, and release.
- 02
Create an App Run
Send the app ID, project ID, and inputJson to the App Runs endpoint. Save the returned run ID.
- 03
Retrieve the output
Poll the run status with backoff. When it reaches a terminal state, collect outputs or inspect errors.
GET
/openapi/v1/apps/{appId}Read the app’s input contract
POST
/openapi/v1/app-runsCreate an asynchronous run
GET
/openapi/v1/app-runs/{runId}Check status and retrieve outputs
Keep your API key in a server-side environment variable.
Authentication & full examplesSTART ON THE CANVAS
Build it once.
Call it whenever you need it.
Your team can experiment visually before anyone writes integration code. Once the workflow works, copy it as an app and publish the version you want to use.
See Workflows & AppsA FEW MORE DETAILS
Good to know.
Do I need a separate Claude Code or Codex plugin?
No separate plugin is required for the HTTP API approach shown here. Your coding assistant can write and run API requests using the MaxDay documentation and credentials you configure in your development environment.
What do I need before I make a request?
An active MaxDay API key, a project ID, and a published app. Read the app details first: each app declares its own inputs. Keep your key server-side and never include it in browser code.
How do I know when an app has finished?
Creating a run returns a run ID. Query that run to check progress and status. Stop polling at succeeded, partial_success, failed, or canceled, and inspect the outputs or error details.
Can I safely repeat the same creation request?
Do not automatically retry a POST that creates an App Run. It may create another run. Use the returned run ID to check the existing run; retry status queries with backoff when needed.
MAKE IT PART OF YOUR NEXT CAMPAIGN