citation-crosscheck

Files (2)

• SKILL.md
• cx_verify.py · scripts/

SKILL.md (first lines)

# Citation Cross-Check (double-blind bib-vs-web field audit)

> 🔒 **Do not wrap this skill in `/loop`, `/schedule`, or `CronCreate`.** It is
> verdict-bearing — it emits a per-reference verdict that gates `.bib` edits (an evidence
> instrument, not a cross-model acquittal; see the acceptance-gate note in Step 2).
> Re-running it on a timer adds no signal (the verdict changes only when the
> *bibliography* changes).
> Schedule the external wait that precedes it: bibliography finalized → run once. See
> [`shared-references/external-cadence.md`](../shared-references/external-cadence.md).

## Why this exists (and how it differs from `citation-audit`)

`citation-audit` is a single cross-model reviewer that judges three things holistically:
existence, metadata, and **context-fit** (is the cited paper used for a claim it
actually supports?). This skill does **not** do context-fit. It does one thing that a
single reviewer cannot do reliably: an **adversarial, double-blind, field-level diff**
of the local `.bib` entry against an *independently reconstructed* record.

The failure mode it targets is the hand-entry error that survives a plausibility read:
a **dropped or re-ordered author**, a **wrong year**, a **swapped venue**, a **DOI that
404s**. A single reviewer that sees the `.bib` entry tends to anchor on it and rate it
"looks correct." Two agents that never see each other's output cannot anchor — a
discrepancy only survives if *both* independent records agree, and the verifier reports
exactly where they diverge.

In practice it catches dropped/re-ordered authors, wrong years, and bad DOIs that a
single-pass audit rated "looks clean" — while `NEEDS_WEBCHECK` stops it from
false-flagging a legitimate non-arXiv (blog / tech-report) citation as fabricated. Use it
**in addition to** `citation-audit`, not instead: `citation-audit` protects your *review
score* (context-fit); this protects
against the *integrity red flag* of a wrong or fabricated reference (existence + fields).

| | `citation-audit` | `citation-crosscheck` (this) |
|---|---|---|
| Owns | existence + metadata + **context-fit** | **existence + field accuracy**, adversarially |
| Method | one cross-model reviewer | **two blind extractors + a mechanical field-diff** |
| Reads the .tex claims? | yes | no (bibliographic only) |
| Deliverable | KEEP/FIX/REPLACE/REMOVE per entry | **side-by-side .xlsx** + MATCH/MINOR/MISMATCH |
| Best at catching | wrong-context citation | dropped/re-ordered author, wrong year/venue/DOI |

## Constants

- **BLIND = true** — the web agent (B) MUST be prompted with only each paper's title +
  first author, plus its opaque `dedup_key` (a run-local ordinal that carries no
  bibliographic content), and MUST NOT read the `.bib` or any repo file. This is the whole
  point; do not relax it.
- **SOURCES** = arXiv (`/abs/<id>`, `/bibtex/<id>`), DBLP, CrossRef, ACL Anthology,
  NeurIPS/PMLR proceedings, OpenReview. Report what the fetched page shows; never
  fabricate a field.
- **VERDICTS** (assigned by the **main agent only**, in Step 4 — never by a shard) =
  `MATCH` (bibliographically equivalent, cosmetic-only) / `MINOR`
  (formatting or a preprint-vs-published venue, no correction required) / `MISMATCH`
  (real discrepancy: wrong/missing author, wrong year, wrong venue, wrong id) /
  `NEEDS_WEBCHECK` (B could not confirm existence, but the entry is a
  `@misc`/blog/tech-report whose `url` B's index-based search structurally cannot see —
  the main agent must fetch that `url` in Step 4 before any verdict; **never** downgrade
  such an entry straight to `MISMATCH`).
- **SCOPE** = cited-only by default (entries that render in the `.bbl`); `--all` also
  covers uncited `.bib` entries.
- **OUTPUT** = `<paper-dir>/citation_crosscheck_<date>.xlsx` (+ the four intermediate JSONs —
  `cx_A_ourbib`, `cx_B_web`, `cx_C_fielddiff`, and the gate's `cx_gate.json` — kept for audit).

## Workflow

### Step 1 — Resolve the citation set (deterministic, main agent)

```bash
# Search the paper's own source tree, not the whole repo: bundled author-kit /
# template .tex carry demo \cite{...} that inflate the set. Set PAPER_DIR to the dir of
# the main .tex (\documentclass + \bibliography); recurse (cites often live in
# sections/*.tex); exclude vendored/example dirs.
PAPER_DIR="<dir of the paper's main .tex>"
cd "$PAPER_DIR"
python3 - <<'PY'
import re, pathlib
# Skip vendored/example/template trees: bundled author kits carry demo \cite{...}.
# NOTE: this is a NAME match — a dir you *want* audited must not match (rename it or
# narrow SKIP), e.g. `examples_of_results/` would be skipped.
SKIP = re.compile(r'(?i)(^|/)(.*template.*|.*upstream.*|.*example.*|.*vendor.*)(/|$)')
# Verbatim-ish blocks AND inline \verb|...| / \lstinline are code samples, not citations.
VERB = re.compile(r'\\begin\{(verbatim|lstlisting|minted|Verbatim)\}.*?\\end\{\1\}', re.S)
# The optional arg may itself contain brackets: \lstinline[language={[LaTeX]TeX}]|...|
VERBINLINE = re.compile(r'\\(?:verb|lstinline)\*?'
                        r'(?:\[(?:[^\[\]]|\[[^\]]*\])*\])?\s*'
                        r'([^\sA-Za-z])(?:(?!\1).)*\1', re.S)
# url/href/path args may contain a literal % that is NOT a comment. Neutralize them
# BEFORE comment-stripping, or a real \cite later on that line is silently lost.
URLCMD = re.compile(r'\\(?:url|href|path|nolinkurl)\s*\{[^{}]*\}')
COMMENT = re.compile(r'(?m)(?<!\\)%.*$')
# EXPLICIT ALLOWLIST of cite commands. Do NOT build this from a prefix/suffix pattern:
# a Cartesian `(no|paren|auto|...)?cite(s|p|t|...)?` both invents commands (\citecite,
# \parenciteyear -> fake keys from any brace arg) and misses real ones (\footfullcite,
# \volcite, \notecite). Longest-first so \footfullcite wins over \footcite.
# volcite-family: the FIRST brace arg is a volume/page, NOT a cite key.
VOLCITE = {'volcite', 'Volcite', 'pvolcite', 'Pvolcite', 'fvolcite', 'ftvolcite',
           'svolcite', 'tvolcite', 'avolcite',
           'volcites', 'Volcites', 'pvolcites', 'Pvolcites', 'fvolcites', 'svolcites',
           'tvolcites', 'avolcites', 'ftvolcite', 'ftvolcites'}
NAMES = sorted({
    # natbib / LaTeX core (capitalized variants are real natbib commands)
    'cite', 'Cite', 'citep', 'citet', 'citealp', 'citealt', 'citenum', 'citeyear', 'citeyearpar',
    'citeauthor', 'citefullauthor', 'Citep', 'Citet', 'Citealp', 'Citealt', 'Citeauthor',
    # biblatex singular + plural
    'parencite', 'Parencite', 'footcite', 'footcitetext', 'footcitetexts', 'textcite',
    'Textcite', 'smartcite', 'Smartcite', 'supercite', 'autocite', 'Autocite', 'fullcite',
    'footfullcite', 'notecite', 'Notecite', 'pnotecite', 'Pnotecite', 'fnotecite',
    'citetitle', 'citeurl', 'citedate', 'autocites', 'parencites', 'textcites',
    'smartcites', 'supercites', 'footcites', 'cites', 'Cites', 'nocite', 'nocites',
    # aliases + field/list forms + bibentry — real biblatex/natbib families
    'citetalias', 'citepalias', 'Citetalias', 'Citepalias', 'citefield', 'citelist',
    'bibentry', 'fullciter',
    *VOLCITE,
}, key=len, reverse=True)
# (?![A-Za-z]) stops \cite matching the \citeXYZ of an unknown command; the trailing run
# accepts any mix of [opt] and {grp} so \cites{k1}[note]{k2} yields BOTH keys.
CITE = re.compile(r'\\(?:' + '|'.join(map(re.escape, NAMES)) + r')\*?(?![A-Za-z])'
                  r'((?:\s*(?:\[[^\]]*\]|\{(?:[^{}]|\{[^{}]*\})*\}))+)', re.S)
# Strip [optional] args BEFORE reading brace groups, or a braced note like
# \cite[see {Smith} p.3]{k} contributes 'Smith' as a phantom key.

From wanshuiyin/Auto-claude-code-research-in-sleep, MIT.

More from wanshuiyin/Auto-claude-code-research-in-sleep