Data-Driven Story Miner

1Mine our own data for earned-media-worthy stories — the highest-ROI PR asset a company owns.
2
3INPUT: data we could analyze/publish: {{data_inventory}} | industry narratives currently running: {{news_context}} | competitive sensitivity limits: {{limits}}
4
5FOR EACH STORY CANDIDATE (find 5):
61. the finding: stated as a headline a journalist would write ("Remote workers send 40% more messages after 9pm").
72. why now: which running narrative it feeds, confirms, or (better) contradicts.
83. data honesty check: sample size, selection bias, and whether the claim survives scrutiny — kill candidates that don't; a debunked stat costs more than it earns.
94. package: chart concept + methodology note + expert quote angle.
105. sensitivity: what this reveals about our business [APPROVAL NEEDED].
11RANK by (news value × data defensibility).
12
13OUTPUT — ONLY JSON: { "stories": [] }