A saddle-tree loft scripts greet → look up a tree → speak a reply, but the lookup table is empty. What should be named as the problem?
Select an answer to reveal the explanation.
Short Explanation
The script says “look up a tree,” but the table shelf is empty. That is a missing data asset for an agent step—not a fusion-type mix-up.
Full Explanation
When a scripted agent path needs a lookup table and that table is empty, the problem is a missing data asset for an agent step. Naming it that way for the saddle-tree loft is correct: greet → look up a tree → speak a reply cannot proceed without the supporting table. Civic agent scripts fail in production for the same reason—procedure without data assets stalls the public interaction mid-flow.
Blaming wrong fusion type on the greeting audio fails; fusion type does not create lookup tables. Calling it a CLIP pairing failure between two stills fails when the gap is a table, not image–text pairs. Calling it an ACE face-drive bug fails by substituting avatar tooling for a missing structured asset.
Exam caveat: agent orchestration exposes missing supporting data as blockers, not as fusion trivia. Operational check: before enabling a civic agent step that fetches, verify the supporting table/API exists and is populated; block the step until it does. Keep municipal reviewers in the loop whenever modality claims, fusion labels, or agent scripts are promoted from lab notes into the production corpus used by public-facing services. Record the decision, owner, and date beside the dataset manifest.