Counts (verified from git log --diff-filter=A):
- 134 insight cards
- 13 synthesis patterns
- 2 contradictions
- 4 methodology playbooks
- 2 new operator profiles (David Ogilvy + Dave Trott; the v6 workflow proved out a "new-operator from scratch" pipeline)
A long day of depth-first ingestion. Six research-scan batches landed, each picking 5 operators and going deep on each — substantive cards with real evidence, not stubs.
What the scans were
- v1 — 5-operator depth-pilot. Tested the protocol: pick a small set, go deep, write cards with full evidence rather than skim-and-stub. +22 cards / 3 patterns / 1 contradiction / 1 playbook.
- v2 — 5-operator batch 2. +26 cards / 3 patterns / 1 contradiction / 1 playbook.
- v3 — 5-operator batch 3 (GTM-economics cluster). Concentrated theme: pricing economics and growth-loop math. +26 cards / 3 patterns / 1 playbook.
- v4 — 5-operator batch 4. +25 cards / 3 patterns.
- v5 — copywriting cluster batch (5 operators). Concentrated theme: voice, structure, the craft of saying it well. +24 cards / 1 pattern / 1 playbook.
- v6 — new-operator workflow. David Ogilvy + Dave Trott. Proved the pipeline for adding operators who weren't in the corpus before — biographical research, primary-source hunt, then card-writing. +11 cards / +2 new operator profiles.
The scans pushed the corpus past the threshold where convergence patterns started to pop on their own — operators reaching the same conclusions from different angles, captured as synthesis/patterns/.