customer-intel

<!-- Generated from `.claude/skills/_available/product/customer-intel/SKILL.md` by `scripts/generate-agents-skills.py`. Do not edit. --> ## Purpose Aggregate and analyze customer feedback from all sources - meeting notes, person pages, feedback captures - to identify patterns, prioritize pain points, and inform product decisions. ## Evidence, authority, and recovery Treat customer intelligence as an evidence ledger, not as memory or a polished guess. - For every finding, retain a stable source ID (vault path, note ID, or capture ID), source type, source date (when the customer said or wrote it), and the as-of date (when this review read it). Keep the source ID and source date attached when a finding is summarized. - Deduplicate repeated copies of the same evidence for frequency counts, but preserve every source ID and source date in the provenance. Distinguish a copied meeting note from an independent mention; duplicate copies are not extra customer mentions. - Preserve quote fidelity: quoted text must be copied exactly, including wording and meaningful punctuation. Mark omissions or inaudible text explicitly, and label any cleaned-up summary as a paraphrase rather than a quote. - Keep unknown and contradictory evidence visible. If sources disagree, show each claim with its source ID and date and call out the contradiction; do not silently choose one. If the records do not support a count, customer attribution, urgency, trend, or roadmap status, write `unknown` and retur...