kai-funnel-auditlisted

## Objective A sourced diagnosis of where the awareness-to-lead flow breaks, in two layers — awareness (do the collected hooks, messaging, and proof earn attention?) and lead capture (do the collected opt-ins convert that attention into leads?) — ending in a provenance-linted, prioritized fix list where each fix is routed to the skill that owns it. Every finding traces to a collected artifact. A surface with no collected artifact is out of scope, not audited from memory. **Scope boundary vs `/kai-cro`:** `/kai-cro` runs the 5-layer conversion stack (technical, traffic, offer, design, copy) on ONE page or flow, in depth. This skill audits the FULL awareness-to-lead path across surfaces — ads, organic posts, entry pages, opt-ins, lead magnets, phone path — and finds where the flow breaks between them. When a single page needs deep conversion work, flag it and hand off to `/kai-cro`; never re-run the 5-layer stack here. `/kai-audit` and `/kai-seo-audit` are wider still; hand off there when the problem is not the funnel. ## Done when Work type `audit-report` — floor **E3/C4/O1** (`harness/eco-floors.yaml`, `also_covers: funnel-audit`). - **E3** — a named human approved the exact delivered folder, and every quantitative claim resolves to a row in `_data-sources.md` backed by an artifact in `workspace/funnel-audit/data/`. - **C4** — the Kai Data Provenance Rule holds end to end: mode declared, collector run before writing, every number cited, gaps written to `_data-gaps.md`. `