You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Scale the proven workflow from #22 across the rest of the slate. Ship episodes 2–8 to hit the 8-article launch floor; episodes 9–10 are stretch additions if cadence allows.
Each article follows the same recipe: extraction script → LLM draft → author curation → disclosure-policy self-review → merge. Articles ship in batches (one PR per article or per small group), preserving review-ability.
Launch slate (Episode 1 ships in #22; this issue covers 2–8 plus stretch):
Building the warehouse: storage choice and category coverage
Why we excluded sports
The first failed paper replication — structural divergence finding
From peak-fill anchor to external calendar
Adding cluster detection and counterparty HHI
Performance lessons: when to measure, when to extrapolate
Long-running scripts and the heartbeat discipline
(stretch) The open-book P&L: designing statements for the project itself
(stretch) Schema-as-contract: how the LP forced a producer/consumer split
Once the floor is hit, run the launch acceptance battery: LLM install-snippet recheck, staging walkthrough, sign-off.
See plans/slice-11-build-log.md Phase 4 for full details.
Acceptance criteria
At least 8 Build Log articles are live, episode-numbered consecutively from 1
Every article covers a distinct phase of poly-track's evolution (no duplicates, no overlapping scope)
Every article passes the author's disclosure-policy self-review and the result is recorded (commit message or PR description)
No published article contains a wallet ID, an alert payload, or content that could front-run an open position
LLM install-snippet manual recheck: paste the homepage install snippet into Claude and ChatGPT, ask "summarize ogsfrompoly's methodology evolution" — both produce substantively correct answers derived from article content alone
Staging walkthrough of /about, /articles/, and a representative article confirms typography, hero rendering, embedded snippet blocks, and feed pickup all behave as designed
Cloudflare Web Analytics confirms hits are recorded for /about, /articles/, and individual article URLs (plus their .md counterparts) within the documented latency
Parent PRD
#12
What to build
Scale the proven workflow from #22 across the rest of the slate. Ship episodes 2–8 to hit the 8-article launch floor; episodes 9–10 are stretch additions if cadence allows.
Each article follows the same recipe: extraction script → LLM draft → author curation → disclosure-policy self-review → merge. Articles ship in batches (one PR per article or per small group), preserving review-ability.
Launch slate (Episode 1 ships in #22; this issue covers 2–8 plus stretch):
Once the floor is hit, run the launch acceptance battery: LLM install-snippet recheck, staging walkthrough, sign-off.
See
plans/slice-11-build-log.mdPhase 4 for full details.Acceptance criteria
/about,/articles/, and a representative article confirms typography, hero rendering, embedded snippet blocks, and feed pickup all behave as designed/about,/articles/, and individual article URLs (plus their.mdcounterparts) within the documented latencypnpm types && pnpm test && pnpm lint && pnpm knipall pass in CIBlocked by
/aboutmust exist for the staging walkthrough)User stories addressed
Reference by number from PRD #12:
.mdcompanion)