Pubmed Verifier

Batch PubMed citation verifier — detect AI-fabricated references in one click. Five-state verdict: ✅ Correct / ⚠️ Mismatch / 🔶 Partial / ❌ Invalid / ❓ Unkno...

This package appears to implement what it claims: it scans files you point it at, queries PubMed (and optionally Crossref), and stores results in a local SQLite cache (~/.cache/pubmed-verifier/cache.db). Before installing or running it: 1) review scripts/verify_pmids.py (already included) to confirm there are no unexpected network endpoints or file operations beyond the documented cache and report outputs; 2) note the provenance/version mismatches (README/skillhub-meta.json vs registry metadata) — if origin matters, try to confirm the author/source; 3) run it on a non-sensitive sample directory first (or in an isolated environment/container) so it can't accidentally read secrets; 4) if you need to avoid local persistence use the --no-cache option or delete the cache file after use; 5) do not run as root. If you want higher assurance, ask the publisher for a signed release or review the full code to ensure there are no hidden exfiltration paths.

Type: OpenClaw Skill Name: pubmed-verifier Version: 2.1.4 The pubmed-verifier skill is a legitimate utility designed to batch-verify PubMed citations (PMIDs) and detect AI-generated reference hallucinations. The core logic in `scripts/verify_pmids.py` uses standard Python libraries to extract PMIDs from local files, fetch metadata via the official NCBI E-utilities and Crossref APIs, and perform fuzzy matching on titles and authors. It implements responsible API usage with rate-limiting and exponential backoff, and uses a local SQLite database at `~/.cache/pubmed-verifier/cache.db` for performance-enhancing caching. No indicators of data exfiltration, malicious execution, or harmful prompt injection were found.