Why the Sume API hides a schedule's instructions text
GET /v1/actions returns the cron, model, cap and schema of a Sume schedule but not its instructions text. Read and edit instructions in the dashboard.

You list schedules with GET /v1/actions and expect the prompt back. It is not there. The Scheduled page states that the public shape does not include the instructions text, calls that intentional, and says to use the dashboard to read and edit instructions.
What the API does return
- Identity and state:
id(aut_...),object: "action",title,status. - Trigger:
trigger_type,api_trigger_enabled,cronwithexpr,timezone,next_run_at. - Behavior:
model,output_schema,primary_output_key,generation_spend_cap_usd_micros. - Timing and entry point:
last_run_at,created_at,updated_at,invoke_url.
What that means for your process
Instructions persist across runs and are the prompt; data from the caller arrives separately as input, as in Advanced: run a schedule via API. So keep a copy of the instructions in your own repo if you need review history, and treat the dashboard as the place they are applied. A prompt change is made there, and the API cannot create, edit or delete a schedule either.
Empty instructions cannot run; a run request against such a schedule fails with 400. If a run behaves oddly, compare the dashboard text with your copy before you blame the model, then check the receipt's output_error and error fields.
Passing per-run variation
Put what changes in input, up to 64 properties and 2 MiB, and keep what stays the same in the saved instructions. That split is also why a cron-only schedule needs no payload, and why the Create a schedule page describes instructions as the prompt that the Agent executes on each run.
Sources
Related posts
More in Agents
- Run the Sume video agent from your backend with Agent Completions
POST /v1/agent/completions runs the same agent as the Sume Agents chat, with tools and media generation, and returns an async run receipt you poll or webhook.
- Safe automation for AI agents that call paid APIs
Keep agents read-only by default, keep secrets out of logs, and on hosted MCP send an idempotency_key, preview with dry_run, and cap with max_spend_usd.
- Scheduled AI video agent runs: cron, API triggers, and receipts
A Sume schedule is a saved Agents automation that runs on a cron cadence and returns a run receipt. Author it in the dashboard; start and monitor runs by API.
- What is a video agent? How Sume defines and runs one
In Sume's docs, a video agent is a sandbox Agent that composes generation tools into a post-ready video. Brief it in chat, or call it over HTTP.
Written by Sume