AIHub Expose String

AIHubExposeString is the text input of the otavanopisto ComfyUI-aihub-workflow-exposer pack: a STRING output whose value the client app supplies, typically a prompt, a filename, or some other free text your workflow needs. In the ComfyUI web UI it's a boring pass-through that returns whatever you typed into value. That's not a bug - the whole pack is built so external apps drive these fields over a websocket, and this node is what turns "the user typed something in their editor" into a string on your graph.

The part that's actually clever is the length policing. This isn't just a text field; it's a text field with rules the client is expected to honor.

The inputs that matter

• value - the string itself. Default empty. multiline - set true to let the client render a textarea instead of a single line (use it for prompts; nobody wants a one-line negative prompt box).
• minlen / maxlen - length bounds (defaults 0 and 1000). The source throws a ValueError if len(value) violates them, so this isn't advisory; a too-short or too-long string fails the run.
• minlen_expose_id / maxlen_expose_id - the interesting pair. Set one to the id of another exposed string, and your bounds become that string's length, plus/minus the matching *_expose_offset. Practical example: a workflow where the output filename must be derived from a title field - tie maxlen_expose_id to that title's id plus an offset for the extension, and the length limit tracks it automatically. The offsets range −1000 to +1000.
• id / label / tooltip / advanced / index - the standard expose fields. id must be unique per workflow; the client sends the value keyed by it.

Output is one STRING. Wire it into a CLIPTextEncode's text for a prompt, a save node's filename, whatever.

Where people get burned

Three failure modes, all easy to hit on first try. First, leaving minlen at something nonzero - set a minimum of, say, 5 and the client will error on any empty/blank submission. Second, the reverse: some client sends a long prompt and maxlen 1000 slams it. Third, minlen_expose_id pointing at an id that doesn't exist or isn't exposed - the reference silently can't resolve and your bounds fall back oddly. Keep the defaults (0 / 1000) until you actually need constraints.

Install is the pack-standard no-fuss path - the pack ships no requirements.txt and pulls only what ComfyUI itself provides:

cd ComfyUI/custom_nodes
git clone https://github.com/otavanopisto/ComfyUI-aihub-workflow-exposer

Restart, and you're done. One honest caveat before you sink time in: like every node in this pack, AIHubExposeString is only useful when an AIHub client is connected. If you're not building an editor integration, this is a pass-through node that would bore a toaster.