Feature prioritization synthesizer
Every planning cycle, product leadership triangulates sales feature-asks, support pain points, and company strategy by hand, and the loudest deal usually wins. Design an agent that synthesizes those inputs into a decision-ready brief. How does it weigh a million-dollar prospect's ask against a thousand small support complaints, where does it stop and the humans start, and how would you evaluate that its briefs are balanced rather than echoing whoever wrote the most? Pitch it to the chief product officer in two sentences.
audit_brief_source_balance(channels: list[str], evidence_counts: list[int], brief_lines: list[int]) → list[str][["sales","support","strategy"],[5,90,5],[60,30,10]]out["sales=over","support=under","strategy=balanced"][["sales","support","strategy"],[16,12,4],[16,12,4]]out["sales=balanced","support=balanced","strategy=balanced"][["sales","support"],[0,50],[20,20]]out["sales=over","support=under"]Turn the fuzzy goal into a bounded agent task. Say what the agent reads, what it may do on its own versus draft for a human, and what it must never touch. Then earn the trust: how you’d verify it before rollout, and the pitch (the problem, what you built, why it matters).
Agent build: design the agent in plain language (or narrate it). The coach grades the decomposition, the guardrails, the verification plan, and the pitch.