Draft the watchlist flag a servicer can defend
Rung 8 of 10 · Cognitive verifier
1 · Learn the move · Cognitive verifier
A watchlist entry gets read by asset management and, eventually, special servicing. So a flag that asserts a breach the numbers don't clearly show is a flag you can't defend. Paste the latest financials and the covenant thresholds, and have the AI draft the entry through a cognitive verifier: for each potential trigger it restates the threshold as pasted, states the reported figure as pasted, and marks KNOWN-BREACH, POSSIBLE, or NO-TRIGGER. The discipline the whole desk turns on: the AI never declares a breach on its own math, a low DSCR reads 'reported 1.05x per servicer report vs 1.20x covenant. Confirm the calculation, then this is a breach,' and missing or stale data makes the flag POSSIBLE, never a guess. Whether to accelerate is not its call. That routes to special servicing and counsel.
Look at these financials against the covenants and tell me if this loan should be on the watchlist. [paste financials + covenant thresholds]
2 · Your turn. You write the prompt
You're an asset manager on a servicing portfolio and this loan's quarterly package just came in soft. DSCR looks thin, one reserve line is blank, and the rent roll is a quarter stale. You paste the reported figures and the covenant thresholds to draft a watchlist entry that names the specific trigger and holds up when special servicing reads it.
Remember: the AI sees only your prompt, not this page. If the situation isn't in your prompt, it doesn't exist.
Optional. These shape the output when you run your prompt below, not your score.