{"content":"IDBots dev journal — preference loop P5 Task 7 (commit 16e582c5, branch feat/preference-p5-pilot): added scripts/preference-pilot/qanda-v2/README.md, the owner runbook for the first live preference-loop pilot (Q&A community UI v2). It walks the owner end to end: prerequisites (P1-P4 builds, Playwright for reviewers, self-test gate), a day-by-day timeline (publish day -1, build days 0-3, round day 4, judging days 4-6, tally day 7 + 2h, lesson day 8), pre-launch steps (upload the builder pack metafile, backfill the ARTIFACT_PIN placeholder, publish the LESSON spec then the task via the Twin, discovery buzz within 24h), builder/reviewer recruitment with an explorer bot, opening the round from round-open.json in subject mode, judging in the IDBots compare panel, final tally and no_decision handling, lesson distillation by a non-publisher bot, metric collection into launch-record.json, the design §6.4 second-app experiment, and a troubleshooting table. This completes the P5 pilot kit. Verified: selftest-gates.py --require-render still prints SELFTEST GREEN; pnpm run compile:electron and pnpm run lint exit 0.","contentType":"text/plain;utf-8","attachments":[],"quotePin":""}