Short answer: there isn’t a built-in “Are you sure?” confirmation modal in the Prefect UI that you can customize per-parameter. But you have a few solid options depending on whether you want to prevent, warn, or gate runs. Options - Prevent invalid inputs at submit time (UI form validation) - Encode constraints in your flow signature so the UI form enforces them. - Works well for enumerations and basic ranges. - Examples: - Restrict to specific choices with Literal:

from typing import Literal
      from prefect import flow

      @flow
      def my_flow(env: Literal["dev", "staging", "prod"] = "dev"):
          ...

- Add numeric bounds with Pydantic annotations (supported by most JSON schema-aware UIs):

from typing import Annotated
      from pydantic import Field
      from prefect import flow

      @flow
      def my_flow(
          threshold: Annotated[int, Field(ge=0, le=100)] = 50,
      ):
          ...

- Or wrap parameters in a Pydantic model if you prefer:

from pydantic import BaseModel, conint
      from prefect import flow

      class Params(BaseModel):
          threshold: conint(ge=0, le=100)
          mode: Literal["fast", "full"] = "fast"

      @flow
      def my_flow(params: Params):
          ...

- Warn or block at run start (code-level guard) - Add a guard at the top of your flow to check ranges and either: - raise a clear error (hard stop), or - log a warning and continue. - Example:

from prefect import flow, get_run_logger

    @flow
    def my_flow(threshold: int):
        log = get_run_logger()
        if not 0 <= threshold <= 100:
            # Choose one:
            # Hard stop:
            raise ValueError(f"threshold {threshold} is outside 0–100")
            # Or soft warn:
            # log.warning("threshold %s is outside 0–100; proceeding anyway", threshold)
        ...

- Require human approval before execution (closest to “Are you sure?”) - In Prefect Cloud, create an Automation that: - Triggers on new flow runs (or on scheduled/running state), - Filters when certain parameter values are out of range, - Applies a “Require approval” action (and optionally sends a Slack/Email notification). - The run will pause awaiting approval in the UI; approving continues, rejecting cancels. This gives you a human-in-the-loop confirmation step without modifying the UI form. If you can share: - Whether you’re on Prefect Cloud or OSS, - If you want to block invalid values or just warn/require approval, - The kinds of constraints you need (enums, numeric ranges, dependent params), I can give you a tailored snippet and (if on Cloud) click-by-click automation setup.