#CLE Guide
1 messages · Page 1 of 1 (latest)
Hi @supple lichen. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
The knowledge sources you provided don’t mention CLE, Classify nodes, or how to reference a Classify node’s name/category in Agent Builder, so I can’t give a definitive, documented way to do what you’re asking.
What is documented is:
- Agent Builder workflows are composed of nodes (start, agent, logic, data, tool, etc.), and you can pass data between them via typed connections and state variables. [Agent Builder]
- Data nodes like Set state are used to define global variables for use across the workflow, typically by taking some upstream node output and storing it. [Data nodes]
However, there is no explicit description in these docs of:
- A “Classify” node’s internal CLE name
- A “category directory in CLE”
- Any syntax for directly referencing a Classify node’s category metadata in a Set state node
Because of that, I can’t reliably tell you how (or if) you can access the CLE name and category from a Classify node and wire them into a Set state node.
If you haven’t already, the most likely pattern (based on how other nodes work) is:
- Inspect the output schema of your Classify node in Agent Builder’s UI (or via the node’s docs, if available).
- In the Set state node, choose that node’s output as the source and map the relevant output field(s) into a state variable.
But this is an inference from how other nodes behave, not something explicitly documented in the sources you shared.
Given the lack of explicit documentation here, your best next step is to:
- Check the node’s output in the Preview panel in Agent Builder to see what fields it exposes, and
- If that still doesn’t surface the CLE name/category, post on or search the OpenAI community forum, where similar widget/state questions have been answered (for example, there’s an official response about upcoming support for widget state, which suggests some of this is still evolving). [Forum post]
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
@supple lichen Kapa.ai is still learning and improving, please let me know how I did by reacting below.