classify-dm-intent

To classify inbound DM intent for lead routing, feed the DM thread and its history into the classifier to produce a structured FSM state, confidence score, and recommended next steps for triaging leads fast.

An acquisition FSM state is a specific stage within a 13-state finite state machine that maps inbound direct message threads to ensure proper routing and next actions during lead qualification.

Yes, you can process DM threads in batches for intent classification. The classifier supports both per-thread single-thread and batch processing sweeps to map inbound messages into acquisition states.

The classifier handles low confidence and hard disqualification scenarios by enforcing operator-review gates, returning a structured classification with confidence scores to flag messages requiring manual intervention.

You need structured YAML output to automate downstream lead qualification. The classifier generates YAML to drive downstream skills with state transitions, confidence scores, and recommended next steps.