Repository Public Research preview
1# Workflows23## Audit report for an author45Lead with what blocks submission. Keep the registry evidence next to each item so the author can6decide without rerunning anything. `verify-bib --format md` produces the tables; add the7sections below around them.89```markdown10# Reference audit: <manuscript>, <date>1112Checked N entries against Crossref, arXiv and doi.org.13Verified: a. DOI found for entries without one: b. Need a decision: c.1415## 1. Must be resolved before submission16- `key` — retracted (notice: <doi>, <date>). Cited in section 2 as support for <claim>.17- `key` — DOI does not exist, and no record with this title was found.18- `key` — DOI resolves to a different paper: cited "<title>", registered "<title>".1920## 2. Probably typos (confirm)21- `key` — year cited 1996, registered 1986.22- `key` — DOI not found; the same title, author and year are registered as <doi>.2324## 3. Could not be checked25- `key` — book, not in the registries. Please confirm from the copy you used.26- `key` — two similar records (<doi>, <doi>); which one did you read?2728## 4. Proposed additions (no change of meaning)29- DOIs for b entries (diff attached).3031## Not covered by this audit32- Whether each cited work supports the sentence that cites it.33- Page numbers of quotations.34- Works published or retracted after <date>.35```3637Rules for the report:3839- One line per entry, with the citation key the author knows.40- Quote the registry value exactly; do not paraphrase a title.41- Separate "the registry disagrees" from "the registry does not know". They call for different42 actions.43- Do not rank references by quality or citation count unless asked.44- If an `error` status remains after a second run, list it under "could not be checked" with the45 reason.4647## Applying fixes4849Only after the author agrees, and one kind at a time:50511. **Add missing DOIs** from `suggested_doi` of `doi_missing_found` entries. Change nothing else52 in the entry.532. **Replace an entry with registry metadata** (`cite <doi> --style bibtex --key <same key>`)54 when the author confirms the record is the work they mean. Keep the citation key so the text55 still compiles.563. **Remove or replace** a retracted or unfound reference: the author rewrites the sentence. A57 claim left without a source must be flagged in the text, not silently kept.5859After editing, run `verify-bib` again on the changed keys (`--keys a,b`).6061## Checking one claim, not one reference6263A verified reference can still be misused. When a sentence relies on a specific result:64651. `get <doi>` and read the abstract when the registry has one.662. `oa <doi>` for a legal open copy; read the passage.673. If no open copy exists, say that the claim was not checked against the text. Do not infer68 the content of a paper from its title.6970## Literature search procedure7172A search is reproducible when someone else can rerun it and understand why each work is in or73out. Keep a log from the first query.7475| Step | Command | Record |76|---|---|---|77| Seeds | `resolve`, `search --title ... --author ...` | DOIs of the seed set |78| Keyword search | `search "<terms>" --from Y --to Y --type journal-article --limit 50` | query, filters, date, total hits, number screened |79| Preprints | `arxiv 'abs:"<phrase>" AND cat:<category>' --sort date` | query, date, hits |80| Backward snowballing | `references <doi>` for each included work | works added |81| Forward snowballing | `citations <doi> --sort citations`, then `--sort date --from Y` | works added, total citing |82| Authors | `author --orcid <id> --sort date` | works added |83| Stop rule | a full round adds nothing new | number of rounds |8485Practical points:8687- Crossref relevance ranking favours title matches; rerun important queries with88 `--sort citations` to surface the established works, and with `--sort date` for recent ones.89- Forward snowballing through highly cited seeds returns thousands of works. Narrow with90 `--from <year>` and read the most cited and the most recent pages rather than all of them.91- Without an OpenAlex key, plan around the daily budget (`references/sources.md`): do the92 backward pass first (Crossref, free), then spend OpenAlex calls on forward citations of the93 few most central works.94- Deduplicate on DOI. A preprint and its published version have different DOIs: `arxiv` output95 gives `published_doi` when the authors recorded it.96- Before finalising, run `retractions` on the included set.9798What to state in the methods section: the services searched and the date, the exact queries and99filters, the snowballing rounds, inclusion and exclusion criteria, and the counts at each step.100Also state that OpenAlex and Crossref were the only indexes used; databases such as Scopus, Web101of Science, PubMed or EconLit may hold works these miss.102