Audit tracking on a Waniwani MCP app

Read the whole project, inventory every tracking call on every surface, verify each event against the defined taxonomy, map the funnel, and report the gaps. The output is a report, not code changes: what is wrong, what is missing, and exactly where each missing event belongs. To apply fixes afterwards, use the instrument-tracking skill.

Source of truth: fetch the live docs first

Event definitions evolve. Before auditing, fetch these pages and treat them as the latest definitions; where they disagree with this file, the docs win:

• docs.waniwani.ai/sdk/tracking/events: the full event catalog and the definition of a qualified lead
• docs.waniwani.ai/sdk/tracking/instrumentation: placement rules per funnel stage
• docs.waniwani.ai/sdk/tracking/identify: identify semantics
• docs.waniwani.ai/sdk/reference/event-schema: typed property shapes

Also read the taxonomy the installed SDK version actually accepts, since that is what typechecks in the project:

bun -e "import { EVENT_TYPES } from '@waniwani/sdk'; console.log(EVENT_TYPES)"