Bib Validate

Bibliography Validation

LIBRARY-FIRST RULE: ALWAYS cross-reference all .bib keys against Paperpile during validation — via ONE dry-run of scripts/bib/rekey_to_canonical.py (Task-Management; matches the local library JSON in ~5s), NOT per-key paperpile search-library loops or lookup sub-agents. This catches key drift between the .bib file and the reference manager. See the Reference Manager Cross-Reference section.

Output Path

Per rules/review-artefact-routing.md (auto-loads in research projects (path-scoped to paper-*/ and paper/)):

• Source slug: bib-validate
• Write reports to: reviews//bib-validate/.md inside the project (where ` is the paper directory slug, e.g., paper-jtp, paper-philtech`). Path is relative to the research project root, not the Task-Management repo.
• Never at project root (./CRITIC-REPORT.md-style filenames are forbidden — pre-rule layout).
• Idempotency: if today's file exists, append a same-day descriptor ({date}-revision.md, {date}-r2.md, {date}-pre-submission.md) — never overwrite.
• Index update: if reviews/INDEX.md exists, write a one-line entry under "Latest per source" pointing at the new file. Otherwise /review-recap will rebuild the index next time it runs.
• Infrastructure repos (Task-Management, atlas-workspace, etc.): this section does not apply — the path-scoped rule won't load there.

Modes

| Invocation | Behaviour | |---|---| | /bib-validate | Standard: cross-reference \cite{} ↔ .bib, library check, preprint staleness | | /bib-validate --verify-doi | Adds DOI resolution: each @article/@inproceedings entry gets its DOI checked via scholarly scholarly-verify-dois. Bib entries with no DOI or unresolvable DOIs are flagged. Use before submission to catch fabricated/hallucinated entries. | | /bib-validate --fix | Auto-fixes resolvable issues in the project .bib; genuinely-missing entries are staged under .paperpile-import/ with their draft cites marked \CiteTodo{...} (build-blocking) for manual Paperpile import. The Paperpile CLI is read-only — it cannot write to the library, so additions are staged, not auto-inserted. | | Combine: --verify-doi --fix | Verify, then fix unverified entries by re-fetching from Crossref |

When to Use

• Before compiling a final version of a paper
• After adding new citations to check nothing was missed
• When biber/bibtex reports undefined citations
• As part of a pre-submission checklist (pair with /proofread)

When NOT to Use

• Finding new references — use /literature for discovery
• Building a bibliography from scratch — use /literature with .bib generation
• General proofreading — use /proofread (which also flags citation format issues)

Phase 0: Session Log (Suggested)

Bibliography validation with preprint staleness checks can be context-heavy (OpenAlex lookups, web searches for published versions). Before starting, suggest running /session-log to capture prior work as a recovery checkpoint. If the user declines, proceed without it.

Convention

Default bibliography file is references.bib — this is the standard across all projects (per the /latex skill convention). However, the skill also supports:

• Any .bib file found in the same directory as the .tex files being audited
• Embedded bibliographies using \begin{thebibliography} / \bibitem{key} blocks
• Both external and embedded simultaneously (rare but possible)

Bibliography Detection

At the start of validation, detect which bibliography method the project uses:

1. External .bib file (standard)

Look for .bib files in the project directory. Priority order:

1. references.bib (preferred — standard naming convention across all projects)
2. Any other .bib file in the same directory as the .tex files

If multiple .bib files are found, validate all of them and produce a combined report. Note which file each issue belongs to. If a legacy-named .bib file (e.g., paperpile.bib) exists alongside references.bib, flag it as a potential cleanup opportunity (the project may have migrated from Paperpile).

Full validation applies: cross-reference checks and quality checks.

2. Embedded \begin{thebibliography} / \bibitem{key}

Some LaTeX documents define references inline rather than using an external .bib file. Detect by scanning .tex files for \begin{thebibliography}.

Extract keys from \bibitem entries:

• \bibitem{key} — standard form, key is the argument in braces
• \bibitem[label]{key} — optional label form (e.g., \bibitem[Smith et al., 2020]{smith2020}), key is in the second set of braces

Only cross-reference checks apply (missing keys, unused keys, typos). Quality checks (required fields, year, author formatting) are skipped because embedded bibliographies don't have structured metadata.

3. Both (rare)

If a project has both a .bib file and \begin{thebibliography} blocks, validate both:

• Run full validation on the .bib file
• Run cross-reference checks on \bibitem entries
• Merge both key sets when checking for missing citations

Workflow

1. Find files: Locate all .tex files in the project
2. Detect bibliography type: Check for .bib files and/or \begin{thebibliography} blocks
3. Extract citation keys from .tex: Scan for all citation commands
4. Extract entry keys from bibliography source(s):

• External: Parse all @type{key, entries from .bib file(s)
• Embedded: Parse all \bibitem{key} and \bibitem[label]{key} entries

1. Cross-reference: Compare the two sets
2. Quality checks: Validate .bib entry completeness (external only)
3. Produce report: Write results to stdout (or save if requested)

Citation Commands to Scan

Scan .tex files for all of these patterns:

| Command | Example | |---------|---------| | \cite{key} | Basic citation | | \citet{key} | Textual: Author (Year) | | \citep{key} | Parenthetical: (Author, Year) | | \textcite{key} | biblatex textual | | \autocite{key} | biblatex auto | | \parencite{key} | biblatex parenthetical | | \citeauthor{key} | Author name only | | \citeyear{key} | Year only | | \nocite{key} | Include in bibliography without in-text citation |

Also handle multi-key citations: \citep{key1, key2, key3}

Cross-Reference Checks

Self-citation handling (double-blind submissions only)

When --double-blind is set OR when a paper-* directory's vault submission frontmatter shows the venue is double-blind, run a self-citation scan. Default expectation: the author CAN and SHOULD cite their own prior work — keep the real bib entry and cite it in the third person. Anonymity comes from removing the author block, not from blinding the bibliography; a {Anonymous} entry actually advertises the self-citation. See rules/double-blind-self-citation.md and _shared/double-blind-anonymity-checklist.md §P4–P5.

1. Load submission author list. Read ~/vault/submissions/--.md authors: / coauthors: field. If absent, prompt: Submission author surnames (comma-separated):.
2. Identify self-citations. For each .bib entry, parse author = {...}, tokenize by and, extract surnames; mark entries whose authors overlap the submission's as self-citations — these are expected, not violations.
3. Scan the .tex body near each self-cite \cite{} (±200 chars) for FIRST-PERSON voice — "our"/"we previously"/"in earlier work of ours". That first-person voice is the actual de-anonymizer, not the names.
4. Severity:

• Info / no action — self-citation cited in the third person with the real entry kept. This is the correct finished state. Naming the authors ("[Author] and [Collaborator] [2] show X") in the third person is fine.
• Critical — first-person voice near a self-cite → rewrite to third person (the citation and the real entry stay).
• Critical (CFP-conditional) — only if the venue's CFP explicitly requires anonymizing self-citations (some security venues, e.g. CCS) → then also blind the entry. Do not blind by default.

1. Output format (first-person-voice case):

`` ⚠ FIRST-PERSON SELF-REFERENCE (Critical) intro.tex:88 near \cite{smith2026paradox} "in our prior work [12] we showed..." fix: third person — "Prior work [12] showed..." (citation + real entry stay) ` For the CFP-mandated-blinding exception only, additionally anonymize the entry: author = {Anonymized for double-blind review}, title = {Title withheld}, note = {Anonymized self-citation}`.

This is a venue-conditional check, not an unconditional desk-reject gate. Run it with the standard cross-reference checks.

Critical: Missing Entries

Citation keys used in .tex but not defined in the bibliography source (.bib file or \bibitem entries).

These will cause compilation errors.

Warning: Unused Entries

Keys defined in the bibliography source but never cited in any .tex file.

Not errors, but may indicate:

• Forgotten citations (should they be \nocite?)
• Leftover entries from earlier drafts
• Entries intended for a different paper

Warning: Possible Typos (Fuzzy Match)

For each missing key, check if a similar key exists in the bibliography using edit distance:

• Edit distance = 1: Very likely a typo
• Edit distance = 2: Possibly a typo
• Flag these with the suggested correction

Common typo patterns:

• Year off by one: smith2020 vs smith2021
• Missing/extra letter: santanna vs sant'anna vs santana
• Underscore vs camelCase: smith_jones vs smithjones

Reference Manager Cross-Reference

After the disk-based cross-reference, check every .bib key against the Paperpile library with ONE dry-run of the canonical resolver:

python3 "$(cat ~/.config/task-mgmt/path)/scripts/bib/rekey_to_canonical.py" 

Produces the full status table (HEALTHY / DRIFT / SUGGESTED / AMBIGUOUS / NOT_FOUND) in ~5s with no per-key CLI calls. Fallback lookups, status semantics, and staleness caveat: [references/ref-manager-crossref.md](references/ref-manager-crossref.md)

Quality Checks on .bib Entries

These checks apply only to external .bib files. Embedded bibliographies lack structured metadata, so quality checks are skipped for them.

Required Fields by Entry Type

| Entry Type | Required Fields | |-----------|----------------| | @article | author, title, journal, year | | @book | author/editor, title, publisher, year | | @incollection | author, title, booktitle, publisher, year | | @inproceedings | author, title, booktitle, year | | @techreport | author, title, institution, year | | @unpublished | author, title, note, year | | @phdthesis | author, title, school, year |

Year Reasonableness

• Flag entries with year current year + 1
• Flag entries with no year at all

Author Formatting

• Check for inconsistent author formats within the file
• Flag entries where author field contains "and others" or "et al." — this is never valid in BibTeX. All authors must be listed explicitly. Severity: Warning.
• Flag entries with organisation names that might need {{braces}} to prevent splitting

DOI Verification & Fabrication Detection

See [references/doi-verification.md](references/doi-verification.md) for hard rules, resolution protocol, and LLM-hallucination detection.

Preprint Staleness Check

For every entry that looks like a preprint, check whether a peer-reviewed version has since been published. Full detection signals, lookup protocol, and classification: [references/preprint-check.md](references/preprint-check.md)

Severity Levels

| Level | Meaning | |-------|---------| | Critical | Missing entry for a cited key — will cause compilation error | | Warning | Unused entry, possible typo, missing required field | | Info | Year oddity, formatting suggestion, bibliography type note |

Bibliography Output

After validation, offer these actions if applicable:

• Embedded bibliography → offer to create references.bib: If the project uses \begin{thebibliography}, offer to extract the references into a proper references.bib file (one @misc entry per \bibitem, with the full text as a note field). The author can then enrich the entries with proper metadata.
• Non-standard .bib name → offer to rename: If the existing .bib file is not named references.bib, offer to rename it to references.bib and update the \bibliography{} command in the .tex file.

These are offers only — do not make changes without explicit confirmation.

Report Format

Full report template with all sections: [references/report-template.md](references/report-template.md)

Sections: Summary table → Critical (missing entries) → Warning (typos, unused, missing fields, DOI mismatches, stale preprints) → Info (year issues) → Limitations (for embedded bibliographies).

Optional: Metadata Verification

See [references/metadata-verification.md](references/metadata-verification.md) for Paperpile and scholarly CLI lookup workflows.

Deep Verification Mode (Parallel, Disk-Based)

See [references/deep-verify-protocol.md](references/deep-verify-protocol.md) for auto-trigger thresholds (≥40 entries) and sub-agent architecture.

Council Mode (Optional)

For high-stakes submissions. Trigger: "council bib-validate", "thorough bib check". Full details: [references/council-mode.md](references/council-mode.md)

Fix Mode

After producing the validation report, automatically fix resolvable issues. DRIFT (paper in library under a different key) → the verified rekey chain: rekey_to_canonical.py --apply (remaps \cite keys in .tex; --extra-map for confirmed SUGGESTED/AMBIGUOUS cases) → rebuild_paperpile_bib.py (regenerates the .bib from canonical Paperpile exports) → citation_lint.py → latexmk compile check. NOT_FOUND → search, then stage the entry as a .bib under .paperpile-import/ and replace its draft cite with \CiteTodo{slug}{title; authors; year; DOI} (build-blocking until imported); metadata → correct BibTeX. Library additions are staged under .paperpile-import/ for manual import — the Paperpile CLI is read-only (no import command; paperpile write-bib --citekeys only exports entries already in the library).

Full auto-fix actions, post-fix maintenance, and skip conditions: [references/fix-mode.md](references/fix-mode.md)

Quality Scoring

When producing a full validation report, apply numeric quality scoring using the shared framework:

• Framework: [../shared/quality-scoring.md](../shared/quality-scoring.md) — severity tiers, thresholds, verdict rules

Map validation findings to the framework tiers:

• Critical (-15 to -25): Missing entry for a cited key (compilation error)
• Major (-5 to -14): DOI mismatch, likely fabricated reference (LLM-drafted hallucination), stale preprint with published version available, "et al." in author field
• Minor (-1 to -4): Missing optional fields, year oddities, unused entries

Compute the score and include the Score Block in the report after the summary table.

Log to REVIEW-STATE.md (final step)

Write the bib-validate report to reviews//bib-validate/.md (where ` is the paper directory slug, e.g., paper-jtp, paper-philtech; create mkdir -p reviews//bib-validate/ first). Then append a row to the project's REVIEW-STATE.md`:

bash ~/.claude/skills/_shared/review-state-log.sh \
  --check bib-validate \
  --paper "" \
  --verdict "" \
  --score "" \
  --open-issues "/" \
  --report "reviews//bib-validate/.md" \
  --notes "" \
  [--trigger "pre-submission-report|review-cluster"]

• Verdict: PASS if no missing/unused/typo'd citations; ISSUES FOUND otherwise.
• Score: the numeric quality score if computed; — otherwise.
• Open issues: total of (missing + unused + typo'd) at

…

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: flonat
• Source: flonat/flonat-research
• License: MIT

Install and usage instructions live in the source repository linked above.