# M1 shipped — v0.2.0 deployed, League Law ready for your morning review

All seven stages of the overnight plan (A–G) shipped; nothing was cut. Scout now runs the full direction loop: intake → briefing export → research import → per-person review game → results → Scouting Report, with zero AI calls in the app itself (D6 held).

## What you'll find at 7am

**https://scout.bowden.works/projects/3 — "League Law — 2026 redesign"** (I renamed the "League Laq" project you created in the browser rather than duplicating it). Darren is a member, materials are attached (his direction distilled in his own words, the full email thread, current-site captures), and **five draft options** imported through the real contract — **nothing published, nobody notified**:

| Option | Source | Pole it tests |
|---|---|---|
| A | Align Family Law (Victoria BC) | Warm & team-first — his "league of teams," made visual |
| B | Reaves & Lee (Dallas) | Approachable & reassuring — practice buttons at the fold |
| C | ELCMD (Maryland) | Crisp & direct — phone bar at the very top, numbered practice zones |
| D | Burnham Law (Colorado) | Bold & confident — his "a little different" ceiling-test |
| E | Podhurst Orseck (Miami) | Polished & authoritative — modern-classic gravitas |

Each has a `why_selected` traced to Darren's quoted criteria. I rejected eight other candidates — notably Avize (direct Victoria PI competitor), and Bick/Roulston/HagEstad for having exactly the dramatic heroes Darren said he hates. **Your 10-minute checklist is the top entry in `.logs/handoff.md`**: review descriptors in the Options tab, optionally walk "Preview the review," publish, then send Darren the link — he asked for exactly this ("send me a couple of examples… I can rank them").

## Verification (numbers, not adjectives)

- **36/36** live-HTTP checks of the whole loop (import validation incl. zip-slip, draft invisibility, 404-not-403, anonymization, autosave idempotence, re-import preserving reviews, results, report, export)
- **42/42** pytest green in-container against real Postgres; `tsc` green in the build (it caught one real seam break, which is its job)
- v0.2.0 live, migration ran, `HEAD /` 405 fixed, security audit: **0 critical, no scout findings**
- Capture tool verified on real sites (14/14), and the one honest gap is recorded in `brief.md`: **no browser-level test of the new UI yet** — API and types only, so your preview walk tomorrow is the first human pass

Built with 5 parallel subagents (review game, admin tabs, results, capture tool, tests) while I kept the schema, contracts, auth boundaries, deploys, and the League Law research. Three substrate frictions logged in `.logs/v8-feedback.md` for the standards library.