- Home
- Skills
- Research & Analysis
- Citation Guard AI Reference Verification & Retraction Check
Works with the AI Agents you already use
Citation Guard AI Reference Verification & Retraction Check
Audit bibliographies against Crossref and PubMed to verify DOIs, flag retractions, and identify metadata mismatches.
$29.99
Citation Guard AI Reference Verification & Retraction Check
Example session with this skill installed
Audit these citations for retractions and metadata accuracy:
- 10.1038/s41586-021-03616-x
- Smith et al., 2023, 'Neural networks in clinical diagnosis', Journal of Tech.
- Read your context and instructions
- Compiled the citation guard ai
I have audited the 2 citations provided against Crossref and PubMed. One record is a verified match, while the other requires manual review due to missing database records.
Citation Audit Report
| Citation ID | Original citation | Primary status | Flags | Fields checked | Evidence source | Evidence URL | Recommended action | Confidence |
|---|---|---|---|---|---|---|---|---|
| 1 | 10.1038/s41586-021-03616-x | VERIFIED_MATCH | None | DOI, Title, Authors, Year | Crossref | https://doi.org/10.1038/s41586-021-03616-x | None | High |
| 2 | Smith et al., 2023, 'Neural networks...' | NOT_FOUND | Missing DOI/PMID | Title, Author, Year | Crossref, PubMed | N/A | Search for DOI or PMID manually | Low |
Executive Summary
- Verified Match: 1
- Not Found: 1
- Retractions/Corrections: 0
Audit Notes
- Citation 1 (DOI: 10.1038/s41586-021-03616-x) matches "Structure and function of a chimeric antimony-binding protein" published in Nature (2021). Metadata is consistent.
- Citation 2 yielded no exact matches in Crossref or PubMed for the title "Neural networks in clinical diagnosis" by Smith in 2023. This may be due to the paper being too recent for indexing or a generic title match failure.
Next steps
- Provide a DOI for the Smith et al. (2023) paper to re-run the audit.
- Confirm the journal name for Citation 2, as "Journal of Tech" may be an abbreviation not recognized by the search API.
- Download the verified metadata for Citation 1 to update your bibliography manager.
Connects securely to your tools. The creator never sees your data.
About this skill
Academic integrity relies on verifiable sources, yet manual citation auditing is slow and prone to oversight. Citation Guard automates the verification of bibliographies by cross-referencing user-supplied data against live Crossref and PubMed records to identify discrepancies, retractions, and missing metadata.
What it does
- Source verification checks DOIs and PMIDs against authoritative scholarly databases to confirm record existence.
- Metadata auditing compares titles, authors, and publication venues to identify field-level mismatches or formatting errors.
- Retraction flagging detects official retraction or correction signals directly from Crossref and PubMed metadata.
- Evidence logging generates a transparent trail including retrieval dates, query strings, and source URLs for every check.
- Priority reporting categorizes citations by status, from verified matches to high-priority mismatches requiring human review.
How it works
- Input the citation list or bibliography and specify a preferred citation style if formatting is required.
- Normalize the data by extracting DOIs, PMIDs, and titles while preserving the original source text.
- Query official Crossref and PubMed API endpoints using identifiers or title-based search logic.
- Audit the retrieved metadata against the input fields to assign a primary status like
VERIFIED_MATCHorRETRACTION_SIGNAL. - Review the generated audit table to address flags, correct metadata, or investigate retracted works.
Frameworks & tools
This skill utilizes the Crossref REST API and NCBI E-utilities (PubMed) for real-time bibliographic data retrieval.
Why this beats prompting it yourself
Generic LLMs often hallucinate DOIs or claim retracted papers are valid because their training data is static. This skill forces live lookups against official registries, ensuring that verification is based on current evidence rather than probabilistic text generation.
Use cases
- Audit a manuscript bibliography before journal submission to ensure all DOIs are functional.
- Screen a literature review for retracted or corrected studies to maintain research quality.
- Clean up inconsistent metadata in a shared reference library or Zotero export.
- Verify the existence of obscure citations that appear potentially fabricated or incorrect.
Known limitations
Does not prove a paper is fabricated if not found. Requires live API access to Crossref and PubMed. Limited to scholarly works indexed in these two databases.
Requires live API access. Does not verify papers not indexed in Crossref or PubMed. Cannot confirm if a paper is 'unretracted', only if a retraction signal exists.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 24 days ago
- Accounts
Crossref
Doi
- Software
Python
Permissions
Allowed Hosts