Skip to content

Latest commit

 

History

History
77 lines (63 loc) · 6.54 KB

next-audit.md

File metadata and controls

77 lines (63 loc) · 6.54 KB

NEXT.md migration audit — 2026-09-12

This is dated evidence, not a live queue. Current status and detailed work instructions live in Beads. Start with NEXT.md and the handoff. Following the migration, the owner requested retaining NEXT.md as a reusable session prompt that selects ready Beads; its former task list remains migrated below.

Completed versus remaining

Work Evidence inspected Verdict
Keeling story site/stories/keeling-curve.md, outline; closed GitHub #9 Draft done; chart port and author voice remain
Planetary Boundaries story prose, outline, chart builder and SVGs; closed datapressr-tzh records 16380d8 and 62c526e Draft done; GitHub #4 stale; author voice remains
Craft/voice research docs/story-craft.md, docs/voice-guide.md; closed datapressr-10y Reuse existing work
Near-term charting docs/charting.md dated 2026-09-06; pinned Plot/jsdom dependencies Observable Plot → SVG decided; native renderer research remains
Enrich first rep co2-ppm/enrich.ts, SUMMARY.md, enriched metadata Done; oil-prices rep remains
Story/enrich skills both drafts have Open questions; neither has .claude/skills symlink Drafted, not activated
Benchmark round 1 docs/structure-benchmark.md; closed datapressr-pao, dlc, jbz, 31i, ejh; helper tests Three reps and eight findings applied; JSON/API and relational join remain
CO₂ extensions typed global annual and growth resources; closed datapressr-7yo Delivered; NOAA growth explicitly differs from annual-mean differences
Drawdown relocation closed datapressr-61n; dataset removed locally Previously completed according to records; current external publication/license comparison remains
Site and six portable skills site pages, six symlinks, closed GitHub #5 Existing deliverables present; live deployment not revalidated

Fresh baseline: npm test passed 28/28; oil-prices and co2-ppm validators each returned zero errors/warnings. These checks do not prove data-level type correctness. Oil-prices deliberately retains Date/Price for comparison with its published reference. Its view says Brent vs WTI but references only Brent; the enrichment task covers this misleading label.

Existing datapressr-gy3 is preserved as the benchmark epic and returned from stale in_progress to open. The original Tesla idea datapressr-03u and all ten closed tasks are preserved. After the migration, the owner selected Tesla as an autonomous discovery/scraping example: datapressr-03u is now a P2 epic, with execution task datapressr-03u.1 covering source discovery, snapshots, reproducible extraction and an autonomy report. No completed implementation was duplicated.

Removed queue coverage

NEXT.md item Canonical Beads
Story #3 trial and activation datapressr-7mc, datapressr-0cp, datapressr-ogt, datapressr-1li, datapressr-9qc
Oil-prices enrich trial and activation datapressr-q96, datapressr-9qc
Benchmark round 2 datapressr-gy3; datapressr-cq6, datapressr-rvi, datapressr-rlb
Keeling chart port datapressr-8rk
Two Drawdown datasets datapressr-jh6, datapressr-aw1
Author voice #1/#2 datapressr-77e
Legacy inbox triage datapressr-h7d; Tesla remains datapressr-03u
Optional site rename datapressr-23l
Data-level validation datapressr-ck8
Docs/changelog publication datapressr-5yq
Monitor/cloud datapressr-46c
v1 completion gate datapressr-9up, datapressr-blj

The referenced GitHub roadmap also preserves native-chart research (datapressr-7fs) and skill evals (datapressr-d6n) as post-v1. datapressr-u3m prepares reviewable corrections to stale GitHub issue bodies without sending messages or changing external status.

Delegation index

The IDs below are a migration snapshot; use Beads for live status, file ownership, steps and acceptance checks.

Bead Deliverable
datapressr-9up DataPressr v1: prove and activate story + enrich
datapressr-q96 Enrich oil-prices and record the second-case findings
datapressr-7mc Story #3: oil-prices argument, source checks, and outline
datapressr-0cp Review story #3 outline and approve the argument
datapressr-ogt Story #3: deterministic Observable Plot charts
datapressr-1li Story #3: write prose from approved outline and charts
datapressr-9qc Resolve draft questions and activate story + enrich
datapressr-blj Review v1 acceptance and publish the completion record
datapressr-cq6 Benchmark round 2: one JSON/REST API source
datapressr-rvi Benchmark round 2: genuine relational multi-file join
datapressr-rlb Benchmark round 2: synthesize findings and scope fixes
datapressr-8rk Port Keeling charts to Observable Plot and polish annotations
datapressr-jh6 Compare both Project Drawdown packages and prepare a decision
datapressr-aw1 Owner decision: Project Drawdown naming and publication
datapressr-77e Author voice pass on Keeling and Planetary Boundaries
datapressr-h7d Prepare evidence and owner choices for the seven legacy inbox finds
datapressr-23l Owner decision: retain or rename the Flowershow site
datapressr-ck8 Design data-level schema validation with executable acceptance fixtures
datapressr-5yq Design publishing docs and changelog on the site together
datapressr-46c Research monitor execution and propose one minimal pilot
datapressr-7fs Research native DataHub/Flowershow chart support after v1
datapressr-d6n Design repeatable evals for prompt-only skills
datapressr-u3m Reconcile legacy GitHub issue status with the Beads audit

Human decisions and post-v1 research are deferred, outside the immediate queue. Drawdown comparison and inbox research prepare concrete choices before owner review. v1 dependencies are encoded explicitly; benchmarks are nonblocking.

Planning defaults are oil-prices for story #3 and an independent AI outline reviewer; author voice remains human. Revise Beads if the owner answers differently. Source selection in benchmark tasks and future architecture choices are explicitly research deliverables, not invented settled implementation specs.

GitHub issues #2–#14 were inspected read-only with gh. No external issues were edited. Closed Beads plus repository artifacts take precedence over stale issue checkboxes. External Drawdown packages and the live deployment still need current verification in their scoped tasks. The former NEXT.md task list and old unattended brief remain recoverable through Git history.