bibverify — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited bibverify (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Use this skill when the user asks to verify a .bib file, repair BibTeX metadata, generate BibTeX from a DOI, explain why a reference lookup source was chosen, or compare original vs updated BibTeX fields.
If the Bibverify MCP server is available, call its tools directly:
doi_to_bibtex: Convert a DOI, DOI URL, or DOI-prefixed value to BibTeX.rank_lookup_sources: Explain the effective lookup order for a title and optional BibTeX entry.explain_update_diff: Compare original and updated BibTeX-like entry objects.verify_bib_file: Verify a .bib file through a Bibverify config file.Prefer doi_to_bibtex for one DOI. Prefer verify_bib_file for project-level reference checks.
If MCP is not configured but the bibverify command is available:
bibverify --doi 10.1038/nature12373 --key example2013
bibverify config.jsonFor first-time setup, tell the user to install Bibverify from PyPI and create a minimal config:
pip install -U bibverify{
"language": "EN",
"bib_file": "references.bib",
"user_info": {
"email": "[email protected]",
"app_name": "Bibverify"
}
}.bib file..bib filename stem. Only an explicit --apply changes the source.not_found result from source_unavailable, and never describe an unindexed reference as fake or fabricated without independent evidence.Summarize what Bibverify checked, which sources were used or preferred, which entries changed, and where generated files were written. For user-facing explanations, focus on DOI-first lookup, dynamic source ranking, field differences, and remaining entries that need manual review.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.