research-repolisted

# Research Repository Scaffold & Audit ## Scope and organizing principle This skill sets up a new research repository, or audits an existing one, around a single organizing principle: **the source library is the spine of the project, and everything else grows from it.** `sources/og/` holds the original PDFs and documents — archival, **gitignored** (size + copyright), rarely read directly. `sources/md/` holds their Markdown conversions, **tracked in git**: this is the LLM-readable knowledge base, and *sources are always read from here, not from the PDFs*. `sources/unprocessed/` is the drop zone where new PDFs land until they are named, converted, and filed. `sources/references.bib` keys the corpus to the manuscript — one entry per source, author+year resolvable to its `md/` file, so a `\cite` key maps to a document actually read. A conversion script (`scripts/convert-sources.sh`, OpenDataLoader PDF) and a `process-source` intake skill keep the three in sync. Get that spine wrong — PDFs scattered, no Markdown, a bib that does not match what was read — and every downstream skill (`literature-review`, `citation-check`, `fact-check`, `paper-tex`) is working on sand. Everything else — `data/`, analysis `scripts/`, `manuscript/` or `paper/`, `review/` + `codebook/`, `figures/`, `tables/`, `replication/` — grows outward from that spine, conditioned on what kind of project this is. ## Instructions ### Step 1. Resolve the target and decide the mode Use the path supplied with the