Update: From doing more research and reading other peoples problems with this - I think I need to find a different platform to achieve my goals. But if the issue below can be resolved let me know.
Message to MindStudio Support / Engineer
I am building a step-by-step “decision safety check” in MindSpring. It takes a proposed action, runs it through a fixed set of checks, saves the results so each step can use them, and then produces a clear report that says whether the action can proceed, must pause, needs escalation, or must be denied, along with the reasons.
I need to confirm whether MindStudio supports a basic but required data-flow pattern for the system I’m building.
I’m implementing a constitutional workflow where:
decisions are evaluated in discrete steps,
each step must deterministically persist its result,
and downstream steps must read that state without recomputation or hidden logic.
The core pattern I need is:
runFunction → persist variable → downstream block reads it
Concretely:
A
runFunctioncomputes or transforms a structured object (for examplegateResults)The function calls
setVariable()to overwrite that variableA later Display or logic block reads
{{gateResults}}reliablyAt the moment, values written with
setVariable()are rendering asundefineddownstream.I understand that MindStudio may require variables to be declared in an output / variables contract, but it’s unclear:
Where in the UI that contract is defined (function asset vs. canvas node)
Whether this pattern is fully supported in CloudRunner at all
Because this is a constitutional workflow, I need:
explicit state persistence,
predictable scope,
and verifiable step-by-step execution.
My questions are:
Is it possible in MindStudio for a
runFunctionto reliably overwrite a workflow variable so downstream blocks can read it?If yes, where exactly do I declare the output variable contract that enables this (panel name / setting location), with a minimal working example?
Clarifying this will determine whether this workflow can be built on MindStudio at all.
Sample code I am trying to run:
Send this exact JavaScript snippet and explain that it runs in a runFunction block.
async function main(config) { // overwrite a workflow variable setVariable("gateResults", { outcome: "PROCEED", rationale_display: "No constraint violation identified." }); }
Then say the downstream Display Content renders:
Outcome: {{gateResults.outcome}} Rationale: {{gateResults.rationale_display}}
sample build:
https://app.mindstudio.ai/agents/gate-result-checker-9782c851/remix
Why this code is ideal
-
It uses no inputs
-
It uses one variable
-
It uses setVariable() exactly once
-
It tests nested object access
-
It matches your constitutional workflow need: compute → persist → render
One-line framing to include
“This is the smallest example of the pattern we need to support: a
runFunctionoverwritesgateResults, and downstream blocks read it deterministically.”
