Skip to content
verify mcp Beta VerifyMCP is currently in beta. If you notice any issues, get in touch and we’ll put it right.

Panda Research Tools

PYPI · PANDA-RESEARCH-TOOLS

Scholarly integrity — citations, literature/PubMed, experts, PDF tables, Wayback, trends (8 tools).

−12 this week 68 Trust /100

Last scanned (10 hours ago)

Scores higher than 61% of PyPI servers

Category Education and Research

Trust breakdown (7 categories)

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →

Supply Chain Security50
  • Malware scan not yet available for this package.Unverified
  • No known CVEs affecting this package version or its production dependencies.Pass
  • Runs hatchling.build at install time, a recognised build step with no custom scripting around it. View diagnostics → Pass
  • 1 of 31 dependencies flagged as unhealthy. View diagnostics → Partial
Provenance & Transparency97
  • Source repository is publicly reachable at the declared URL. View diagnostics → Pass
  • Cryptographically verified build provenance (signed, bound to pandadad0214/panda-research-tools). View diagnostics → Pass
  • Clear OSI-approved license (MIT).Pass
  • Actively maintained (last published 10 days ago).Pass
  • Disclosure check failed: no security disclosure policy was found in the source repository. See how to fix → Fail
Schema Quality & AI Usability56
  • AI-judged instruction clarity (excellent).Pass
  • Context-footprint check failed: tool/resource definitions use about 9293 tokens (~1161/item across 8 items; 8 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
  • Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management40
  • Stability observed for 12 of 30 days with no destabilising changes; credit accrues until the full window elapses.Partial
Tool Coverage100
  • 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
  • 100% of tool parameters carry a description.Pass
Tool Safety100
  • No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
  • We read all 8 captured tool definition(s), and no name or description among them implies an irreversible operation.Pass
  • An AI judge read all 9 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
  • Implements a current MCP spec version (2026-07-28).Pass
What we tested

Checks we ran (15)

  • Known CVEs
  • Install scripts
  • Dependency health
  • Source repository
  • Provenance
  • License
  • Maintenance
  • Security disclosure
  • Schema quality
  • Stability
  • Tool coverage
  • Injection markers
  • Destructive annotations
  • Judged manipulation
  • MCP protocol

Checks we could not complete (1)

These are gaps on our side, not findings about this package. We only credit what we can confirm, so a category made up entirely of them scores 0.

  • Malware scan in Supply Chain SecurityMalware scan not yet available for this package.

Does not apply to this package (2)

  • Endpoint SecurityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
  • Transport & ReachabilityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.

The score does not measure everything. What the score does not measure →

Changelog

Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.

  • 21 Nov 26 −15
    • Malware scan: pass → unverified ▼ security
  • 20 Nov 26 +1

    No change was recorded against any check on this day. Stability & Change Management went from 33 to 37. That category is still filling its 30-day observation window: 10 days of observed history at the previous scan, 11 at this one. The score rises as the window fills, whether or not the server changes.

  • 18 Nov 26 +16
    • Malware scan: unverified → pass ▲ security
  • 17 Nov 26 −15
    • Malware scan: pass → unverified ▼ security
  • 16 Nov 26 +1

    No change was recorded against any check on this day. Stability & Change Management went from 20 to 23. That category is still filling its 30-day observation window: 6 days of observed history at the previous scan, 7 at this one. The score rises as the window fills, whether or not the server changes.

  • 14 Nov 26 +1

    No change was recorded against any check on this day. Stability & Change Management went from 13 to 17. That category is still filling its 30-day observation window: 4 days of observed history at the previous scan, 5 at this one. The score rises as the window fills, whether or not the server changes.

  • 12 Nov 26 +16
    • Malware scan: unverified → pass ▲ security
  • 10 Nov 26 +1
    • Schema quality: 8345 → 9293 ▼ functional
    • Stability: unverified → 0.03 ▲ functional
    • Package version: 0.1.0 → 0.2.0 functional
Diagnostics

Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.

Captured 21 Nov 2026 · Analysed pypi/panda-research-tools@0.2.0

Provenance Verified

A signed build attestation was found and verified, binding this exact artifact to the source repository it claims to come from.

Result Verified
Ecosystem pypi
Reason Verified
Discovered via Registry attestation endpoint
Source repo pandadad0214/panda-research-tools
Certificate issuer https://token.actions.githubusercontent.com
Certificate SAN https://github.com/pandadad0214/panda-research-tools/.github/workflows/release.yml@refs/tags/v0.2.0
Rekor log index 3789097861
Predicate type PyPI publish attestation https://docs.pypi.org/attestations/publish/v1
Subject digest sha256:5b6f85761797e445a40572cbacaf82c9e858214f66a764f84bd1dcfe4a364455

Background: How many MCP packages publish verified provenance →

Install scripts 1 script
Hook Tier Command
build_backend Known build tool hatchling.build

Background: Why install scripts are a supply-chain risk →

Dependencies 31 packages
Packages resolved 31
No linked repository 1
Tree resolution Complete

Background: SBOMs and build attestations, explained →

MCP tools · 8 exposed · ~9,066 tokens

The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →

Tool Tokens
citation_formatter ~1,091

Convert DOIs, ISBNs, PMIDs, arXiv IDs, titles or messy reference strings into checked citations in APA 7, MLA 9, Chicago, Harvard, IEEE, Vancouver, BibTeX and RIS. Unlike reference_checker, it formats rather than audits: unresolved items are listed, not diagnosed as fake or retracted. Use when you need a clean bibliography or .bib/.ris export. Do not use to find papers (use literature_search) or to prove references are genuine (use reference_checker first). Parameters: pass references (one per line) or paste a whole list or .bib/.ris contents in text. All styles are one price, so request every style you might need in one call rather than calling again. Text lookups below minMatchScore stay unresolved; DOIs and IDs resolve exactly. Only formatted citations are billed. Returns per item: summary, status, one field per style (apa, mla, ...), inText, bibtex, ris, doi, title, authors, year, containerTitle; combined bibliography files stay in the run's Apify key-value store. Behavior: read-only; runs the Apify Actor pandadad0214/citation-formatter (https://apify.com/pandadad0214/citation-formatter) on your own Apify account. Price: $0.003 per citation formatted. Runs that find nothing are not charged per result, and spending is capped per call by run_max_charge_usd (default $0.50). run_max_items (default 50) limits items returned and fills maxReferences when you leave it unset, so it also limits what is billed; setting maxReferences higher can bill more than is returned. Most calls finish in 5-15 s for a few inputs; large inputs or crawls can take minutes, so raise run_timeout_secs (max 300) or shrink the input after a timeout error. Returns JSON {actor, count, items}; long text output is trimmed to whole items, with all items in structuredContent.

NameTypeReqDescription
contactEmailstring–Contact email for Crossref (optional). Optional. Sent to Crossref and NCBI as the contact so requests use their faster 'polite' pools. Never shared elsewhere.
fixAllCapsTitlesboolean–Fix ALL-CAPS titles. Some registry records store titles in capitals. Convert them to sentence case. Default: on.
includeBibtexboolean–BibTeX export. Add a BibTeX entry per reference and a combined .bib file (LaTeX, Overleaf, JabRef). Default: on.
includeRisboolean–RIS export. Add an RIS entry per reference and a combined .ris file (Zotero, Mendeley, EndNote, Rayyan). Default: on.
listUnresolvedInFilesboolean–List unresolved references in the bibliography files. Append references we could not find to the end of each bibliography file, marked [UNRESOLVED - check manually], so nothing silently disappears. T…
maxReferencesinteger–Max references per run. Safety cap. Only formatted citations are billed. Whole number, 1 to 20,000 references; default 1,000.
minMatchScoreinteger–Minimum match score for text lookups (50-100). How closely a search result must match a free-text reference (title, authors, year) to be used. DOIs and ISBNs are exact lookups and are not affected. W…
referencesarray–References (one per line). DOIs, DOI links, ISBNs, arXiv IDs, PMIDs ("PMID: 123"), paper titles, or messy reference strings in any style. Each one is looked up and formatted.
run_max_charge_usdnumber–Optional. Hard spending cap for this call in USD (default from PANDA_MAX_CHARGE_USD, 0.50).
run_max_itemsinteger–Optional. Most results to return from this call (default from PANDA_MAX_ITEMS, 50). Also limits how many results the Actor produces and charges for, unless you set its own max field.
run_timeout_secsinteger–Optional. Seconds to wait for results (default from PANDA_TIMEOUT_SECS, 120; max 300).
sentenceCaseTitlesboolean–Sentence-case article and book titles. APA, Harvard, IEEE and Vancouver use sentence case ("Deep residual learning for image recognition"). Registries often store Title Case, so we convert it; acrony…
stylesarray–Citation styles. Styles to output. All six are included in the price, so pick as many as you like.
textstring–Paste a reference list or file contents. Optional. Paste a whole bibliography (numbered, one per line, or blank-line separated), or the contents of a .bib (BibTeX) or .ris file. BibTeX/RIS entries ar…

No output schema declared.

No examples provided.

literature_search ~1,352

Search 250M+ scholarly works (OpenAlex) by topic query or seed DOIs and return ranked papers with DOI, citation count, free PDF link, retraction flag and a ready citation. Unlike pubmed_evidence_table (PubMed biomedical records with study design and PRISMA counts), this covers every discipline and supports citation-network expansion. Use when you need real papers on a topic, a reading list, or papers citing/cited by known DOIs. Do not use for systematic biomedical screening (use pubmed_evidence_table) or to find people (use researcher_expert_finder). Parameters: seedDois with no query walks the citation network per seedMode (citing = forward, references = backward); with a query it searches inside that network. searchIn=fullText broadens recall but lowers precision. retractedPapers=flag returns and bills retracted papers marked FIX; exclude drops them unbilled. sortBy changes dataset order only; readingListSize does not affect cost. Returns per item: summary, status, title, authors, year, venue, doi, citations, readingRank, readingScore, isOpenAccess, pdfUrl, retracted, citation, bibtex, ris, abstract. Behavior: read-only; runs the Apify Actor pandadad0214/literature-search (https://apify.com/pandadad0214/literature-search) on your own Apify account. Price: $0.003 per paper found. Runs that find nothing are not charged per result, and spending is capped per call by run_max_charge_usd (default $0.50). run_max_items (default 50) limits items returned and fills maxResults when you leave it unset, so it also limits what is billed; setting maxResults higher can bill more than is returned. Most calls finish in 5-15 s for a few inputs; large inputs or crawls can take minutes, so raise run_timeout_secs (max 300) or shrink the input after a timeout error. Returns JSON {actor, count, items}; long text output is trimmed to whole items, with all items in structuredContent.

NameTypeReqDescription
checkUnpaywallboolean–Find open-access PDFs (Unpaywall). Look up the best legal open-access copy of each paper. Default: on.
citationStylestring–Citation style. Style for the citation column and the CITATIONS file. Formatted by citeproc-js with the official CSL style. Choices: apa = APA 7th edition; mla = MLA 9th edition; chicago = Chicago au…
contactEmailstring–Contact email for API etiquette (optional). Sent to Crossref and Unpaywall as the polite-pool contact. Leave empty to use ours.
includeAbstractsboolean–Include abstracts. Add the abstract where OpenAlex has one (some publishers withhold them). Default: on.
languagestring–Language. Optional two-letter language code, e.g. en, de, es.
maxResultsinteger–Maximum papers. You pay per returned paper, so this also caps the cost ($0.003 x papers). Whole number, 1 to 2,000 results; default 100.
minCitationsinteger–Minimum citations. Only return papers cited at least this many times (OpenAlex counts). Whole number, 0 or more citations; default 0.
openAccessOnlyboolean–Open access only. Only return papers with a free, legal full-text copy. Default: off.
querystring–Search query. Topic or Boolean query. Use quotes for phrases and AND / OR / NOT in capitals, e.g. "retrieval augmented generation" AND (clinical OR medicine) NOT survey. Leave empty to explore around…
readingListSizeinteger–Reading list length. How many top papers go into the ranked READING_LIST file. All papers are still in the dataset and exports. Whole number, 5 to 500 papers; default 25.
retractedPapersstring–Retracted papers. Flag keeps retracted papers in the results, marked FIX, so you know to avoid them (they are charged as returned papers). Exclude drops them before they are returned (not charged). C…
run_max_charge_usdnumber–Optional. Hard spending cap for this call in USD (default from PANDA_MAX_CHARGE_USD, 0.50).
run_max_itemsinteger–Optional. Most results to return from this call (default from PANDA_MAX_ITEMS, 50). Also limits how many results the Actor produces and charges for, unless you set its own max field.
run_timeout_secsinteger–Optional. Seconds to wait for results (default from PANDA_TIMEOUT_SECS, 120; max 300).
searchInstring–Search in. Title and abstract is the most precise (like a database search). Full text also matches papers that only mention your terms in the body. Choices: titleAbstract = Title and abstract (recomm…
seedDoisarray–Seed papers (DOIs). Optional. Start from papers you already trust. Combine with a query to search inside that network, or use alone. Up to 50 DOIs.
seedModestring–Seed mode. Which papers around the seeds to return. Choices: citing = Papers that cite the seeds (forward); references = Papers the seeds cite (backward); related = Related papers (OpenAlex similarit…
sortBystring–Sort results by. Order of the dataset. The reading list is always ranked by reading score (relevance plus field-normalised citation impact). Choices: relevance = Relevance; citations = Most cited; ne…
typesarray–Document types. Leave empty for all paper types (peer reviews, errata, retraction notices and abstracts-only records are always left out unless you pick them here).
yearFrominteger–From year. Earliest publication year (inclusive). Whole number, 1500 to 2100.
yearTointeger–To year. Latest publication year (inclusive). Whole number, 1500 to 2100.

No output schema declared.

No examples provided.

pdf_table_extractor ~1,222

Extract every table from PDF files (by URL or from an Apify key-value store) into structured rows, Excel and CSV, with optional OCR for scanned pages. Unlike sec_financials_excel (standardized XBRL financials straight from SEC filings), this reads whatever tables a PDF contains. Use when the data you need is locked in PDF tables: reports, statements, government releases. Do not use for SEC-listed company financials (use sec_financials_excel, which is exact) or to find papers (use literature_search). Parameters: billing is per page analysed plus per table extracted, so pages and maxPagesPerPdf are the real cost levers; run_max_items does not cap pages. ocrScannedPages=false skips and lists scanned pages; turning it on adds the per-OCR-page price up to maxOcrPagesPerPdf, and is much slower. detection=auto tries ruled-grid then whitespace detection; force lattice or stream only if auto merges or splits columns. Files over 150 MB fail and are not charged. Returns per item (one per table): summary, pages, header, row and column counts and cell data; Excel/CSV files stay in the run's Apify key-value store. Behavior: read-only; runs the Apify Actor pandadad0214/pdf-table-extractor (https://apify.com/pandadad0214/pdf-table-extractor) on your own Apify account. Price: $0.001 per page analysed; $0.003 per table extracted; $0.02 per scanned page ocr. Runs that find nothing are not charged per result, and spending is capped per call by run_max_charge_usd (default $0.50). run_max_items (default 50) limits items returned but does not limit what the Actor bills; use the tool's own limits and run_max_charge_usd. Most calls finish in 5-15 s for a few inputs; large inputs or crawls can take minutes, so raise run_timeout_secs (max 300) or shrink the input after a timeout error. Returns JSON {actor, count, items}; long text output is trimmed to whole items, with all items in structuredContent.

NameTypeReqDescription
detectionstring–Table detection. Auto tries ruled-grid detection first and switches to whitespace (column alignment) detection when that recovers more of the table. Pick one mode if you know your PDF's style. Choice…
headerDetectionstring–Header row. Auto finds header rows above the first data row (multi-line headers are joined). First row = always use row 1. None = generic Column 1, Column 2… names. Choices: auto = Auto-detect; first…
maxOcrPagesPerPdfinteger–Maximum OCR pages per PDF. OCR has its own cap because it is slow and priced per page. Extra scanned pages are flagged, not charged. Whole number, 0 to 100 pages; default 10.
maxPagesPerPdfinteger–Maximum pages per PDF. Caps cost and run time on very long documents. You pay per page analysed. Whole number, 1 to 2,000 pages; default 100.
mergeAcrossPagesboolean–Join tables that continue on the next page. When a table carries on over a page break with the same columns, output it as one table with a page range. Default: on.
mergeWrappedRowsboolean–Merge wrapped rows. Join labels that wrap onto a second line back into one row, so numbers stay on the right line. Default: on.
numbersAsNumbersboolean–Numbers as numbers in Excel. Write 1,234 / (56) / 4.2% as real Excel numbers (parentheses = negative) so you can sum and chart them. CSV and dataset keep the text exactly as printed. Default: on.
ocrScannedPagesboolean–OCR scanned pages (beta). Pages with no text layer (scans, photos) are always detected. Off (default): they are skipped, listed in SUMMARY and on the Index sheet, and cost nothing. On: they are read…
pagesstring–Pages. Optional. Only these pages, e.g. 1-5, 8, 12-end. Leave empty for all pages (up to the page limit below).
pdfStorestring–Key-value store with PDFs. Optional. Pick a key-value store in your account that holds PDFs. The Actor gets read access to this store only.
pdfStoreKeysarray–Record keys in that store. Optional. Keys of the PDFs to process (e.g. REPORT.pdf). Leave empty to process every PDF record in the store (up to 50).
pdfUrlsarray–PDF URLs. Direct links to PDF files. Google Drive and Dropbox share links are converted to direct downloads automatically. Up to 150 MB per file.
run_max_charge_usdnumber–Optional. Hard spending cap for this call in USD (default from PANDA_MAX_CHARGE_USD, 0.50).
run_max_itemsinteger–Optional. Most results to return from this call (default from PANDA_MAX_ITEMS, 50). Also limits how many results the Actor produces and charges for, unless you set its own max field.
run_timeout_secsinteger–Optional. Seconds to wait for results (default from PANDA_TIMEOUT_SECS, 120; max 300).

No output schema declared.

No examples provided.

pubmed_evidence_table ~1,227

Run one or more PubMed queries and return a screening-ready evidence table: one row per article with PMID, journal, study design, sample size, retraction status and citation, plus PRISMA-style counts for the whole search. Unlike literature_search (all disciplines, OpenAlex ranking, citation-network walks), this uses native PubMed syntax and MeSH and is built for systematic-review screening. Use when the question is biomedical/clinical and you need reproducible PubMed queries or design filters (RCT, meta-analysis, cohort). Do not use for non-biomedical topics (use literature_search), for trial registrations (use clinical_trial_leads) or to verify a reference list (use reference_checker). Parameters: queries accept PubMed field tags ([tiab], [mh], [pt]) and Boolean operators; several queries are de-duplicated into one table. Filters (studyDesigns, excludeAnimalOnly, hasAbstract, freeFullTextOnly, excludePreprints) run before billing, and PRISMA counts still report what they removed. retractedArticles=flag bills retracted rows marked FIX; exclude drops them unbilled. sampleSize is parsed from the abstract and comes with sampleSizeConfidence; treat low confidence as unverified. Returns per item: summary, status, pmid, title, year, journal, studyDesign, sampleSize, sampleSizeConfidence, retractionStatus, retracted, citation. Excel, PRISMA and RIS/Rayyan import files stay in the run's Apify key-value store. Behavior: read-only; runs the Apify Actor pandadad0214/pubmed-evidence-table (https://apify.com/pandadad0214/pubmed-evidence-table) on your own Apify account. Price: $0.004 per article found. Runs that find nothing are not charged per result, and spending is capped per call by run_max_charge_usd (default $0.50). run_max_items (default 50) limits items returned and fills maxResults when you leave it unset, so it also limits what is billed; setting maxResults higher can bill more than is returned. Most calls finish in 5-15 s for a few inputs; large inputs or crawls can…

NameTypeReqDescription
citationStylestring–Citation style. Style for the citation column and the CITATIONS file. Formatted by citeproc-js with the official CSL style. Choices: vancouver = Vancouver (NLM); apa = APA 7th edition; ieee = IEEE; h…
contactEmailstring–Contact email for NCBI / Crossref (optional). Sent to NCBI and Crossref as the contact for this tool. Leave empty to use ours.
dateFromstring–Published from. Optional. 2020, 2020-06 or 2020-06-30 (publication date).
dateTostring–Published to. Optional. 2025, 2025-12 or 2025-12-31.
excludeAnimalOnlyboolean–Exclude animal-only studies. Removes records indexed with Animals but not Humans. Records not yet indexed are kept. Default: off.
excludePreprintsboolean–Exclude preprints. Removes records with the Preprint publication type. Default: off.
freeFullTextOnlyboolean–Free full text only. Keep only records PubMed links to a free full text. Default: off.
hasAbstractboolean–Only records with an abstract. Removes records without an abstract (often letters, editorials and older records). Default: off.
includeAbstractsboolean–Include abstracts. Include the abstract in the table and import files (recommended for screening). Default: on.
languagesarray–Languages. Optional, e.g. english, spanish. Leave empty for all.
maxResultsinteger–Maximum articles. You pay per returned article ($0.004 each), so this also caps the cost. Counts for everything not included are still reported. Whole number, 1 to 10,000 results; default 200.
queriesarrayyesPubMed search strings. One or more PubMed queries, written exactly as you would in PubMed (field tags like [tiab], [mh], [pt] and AND / OR / NOT work). Records found by more than one query are counte…
retractedArticlesstring–Retracted articles. Flag keeps them in the table, marked FIX (charged as returned articles). Exclude removes them before they are returned (not charged) and counts them in PRISMA. Choices: flag = Fla…
run_max_charge_usdnumber–Optional. Hard spending cap for this call in USD (default from PANDA_MAX_CHARGE_USD, 0.50).
run_max_itemsinteger–Optional. Most results to return from this call (default from PANDA_MAX_ITEMS, 50). Also limits how many results the Actor produces and charges for, unless you set its own max field.
run_timeout_secsinteger–Optional. Seconds to wait for results (default from PANDA_TIMEOUT_SECS, 120; max 300).
sortBystring–Order. Order of the table. PubMed best match or most recent first. Choices: relevance = Best match (PubMed relevance); newest = Most recent.
studyDesignsarray–Study designs. Keep only these designs (PubMed publication types / MeSH). Leave empty for all.

No output schema declared.

No examples provided.

reference_checker ~1,037

Verify an existing bibliography against Crossref: flags fake (AI-hallucinated), miscited and retracted references, one row per reference. Unlike citation_formatter, it judges whether a reference is real and accurate; it does not produce formatted styles. Use when you must audit a reference list, DOIs or titles before submission or review. Do not use to discover new papers (use literature_search or pubmed_evidence_table) or to restyle citations (use citation_formatter). Parameters: give bibliography, dois or titles (at least one). Lowering minMatchScore finds more but risks false matches; includeCandidates adds audit detail without extra charge. mode=monitor re-checks the DOIs/ORCID works against the state saved under monitorName and reports only new problems, so reuse the same monitorName on scheduled runs. Each reference checked is billed, so maxReferences/run_max_items cap cost on long lists. Returns per item: summary, status (OK/CHECK/FIX), verdict, matchedDoi, matchScore, mismatches (cited vs actual), integrity.status (retracted, corrected, expression-of-concern), openAccessUrl, warnings. Behavior: read-only; runs the Apify Actor pandadad0214/reference-checker (https://apify.com/pandadad0214/reference-checker) on your own Apify account. Price: $0.01 per reference checked. Runs that find nothing are not charged per result, and spending is capped per call by run_max_charge_usd (default $0.50). run_max_items (default 50) limits items returned and fills maxReferences when you leave it unset, so it also limits what is billed; setting maxReferences higher can bill more than is returned. Most calls finish in 5-15 s for a few inputs; large inputs or crawls can take minutes, so raise run_timeout_secs (max 300) or shrink the input after a timeout error. Returns JSON {actor, count, items}; long text output is trimmed to whole items, with all items in structuredContent.

NameTypeReqDescription
bibliographystring–Bibliography / reference list. Paste a reference list in any common style (APA, MLA, Chicago, Vancouver, IEEE, Harvard). One reference per line, numbered ([1], 1.), or separated by blank lines.
checkOpenAccessboolean–Find open-access links. Look up free legal full-text links via OpenAlex. Default: on.
contactEmailstring–Contact email for Crossref (optional). Optional. Sent to Crossref as the mailto contact so requests use its faster 'polite' pool. Never shared elsewhere.
doisarray–DOIs. DOIs to check (bare 10.xxxx/..., doi: prefix or https://doi.org/ URLs).
includeCandidatesboolean–Include candidate matches. Also output the top Crossref candidates and their scores for each reference (useful for auditing). Default: off.
maxReferencesinteger–Max references per run. Safety cap on references checked (each is billed). Whole number, 1 to 20,000 references; default 1,000.
minMatchScoreinteger–Minimum match score (0-100). How close a Crossref record must be (title, authors, year) to count as found. Lower finds more but risks false matches. Whole number, 50 to 100; default 80.
modestring–Mode. 'check' = the normal one-off check. 'monitor' = re-check the DOIs (and/or ORCID works) and compare with the previous run's saved state. Choices: check = Check references (one-off); monitor = Mo…
monitorNamestring–Monitor name (monitor mode). Optional. Name of the key-value store in your account that keeps state between runs (letters, digits, hyphens). Use a different name for each list you monitor. Default: r…
orcidstring–ORCID iD (monitor mode). Optional. Monitor every DOI on this researcher's public ORCID record (plus Crossref works tagged with it), e.g. 0000-0002-1825-0097. Combined with the DOIs list.
run_max_charge_usdnumber–Optional. Hard spending cap for this call in USD (default from PANDA_MAX_CHARGE_USD, 0.50).
run_max_itemsinteger–Optional. Most results to return from this call (default from PANDA_MAX_ITEMS, 50). Also limits how many results the Actor produces and charges for, unless you set its own max field.
run_timeout_secsinteger–Optional. Seconds to wait for results (default from PANDA_TIMEOUT_SECS, 120; max 300).
titlesarray–Paper titles. Article titles to look up when you have no full reference.

No output schema declared.

No examples provided.

researcher_expert_finder ~1,240

Rank the academic experts and key opinion leaders on a research topic from OpenAlex, with institution, ORCID, h-index, topic paper counts and their recent topic papers. Unlike literature_search (returns papers) this returns people; unlike company_team_finder (staff listed on a company website) it covers academics by publication record. Use when you need reviewers, KOLs, collaborators or speakers for a topic, optionally by country or institution. Do not use for company contacts (use company_team_finder) or for papers themselves (use literature_search). Parameters: the ranked pool is the 200 authors with the most matching papers, so maxResults above 200 adds nothing. maxAuthorsPerPaper ignores mega-collaboration papers that would inflate everyone on them; raise it only for physics-style fields. yearFrom/yearTo limit which papers count, while activeSince filters people. includeOrcid adds employment and bio at no extra charge. Returns per item: expert score, name, ORCID, current institution and country, h-index, total and topic works/citations, recent topic papers (title, year, DOI, citations), summary. Behavior: read-only; runs the Apify Actor pandadad0214/researcher-expert-finder (https://apify.com/pandadad0214/researcher-expert-finder) on your own Apify account. Price: $0.0025 per expert found. Runs that find nothing are not charged per result, and spending is capped per call by run_max_charge_usd (default $0.50). run_max_items (default 50) limits items returned and fills maxResults when you leave it unset, so it also limits what is billed; setting maxResults higher can bill more than is returned. Most calls finish in 5-15 s for a few inputs; large inputs or crawls can take minutes, so raise run_timeout_secs (max 300) or shrink the input after a timeout error. Returns JSON {actor, count, items}; long text output is trimmed to whole items, with all items in structuredContent.

NameTypeReqDescription
activeSinceinteger–Active since (year). Optional. Only keep people who published anything in or after this year (uses OpenAlex yearly counts for the last ~10 years). Whole number, 1900 or more.
countriesarray–Countries. Optional 2-letter ISO country codes (US, GB, DE, IN, AU…). Keeps experts whose current institution is in one of these countries, and counts only papers with an author in them.
includeOrcidboolean–Add ORCID details. For experts with an ORCID iD, add current employment and role, keywords, websites and a short biography from the public ORCID record. Default: on.
institutionstring–Institution. Optional. University, hospital, lab or company name (e.g. Stanford University), a ROR ID or an OpenAlex institution ID (I…). Sub-units count too: a university includes its schools and ho…
maxAuthorsPerPaperinteger–Ignore papers with more authors than. Large collaboration papers (hundreds of authors) would make everyone on them look like an expert. Papers with more authors than this are not counted. Whole numbe…
maxResultsinteger–Maximum experts. You pay per expert returned, so this also caps the cost. Up to 200 (the ranked pool is the 200 authors with the most matching papers). Whole number, 1 to 200 results; default 25.
minCitationsinteger–Minimum total citations. Optional. Whole number, 0 or more citations.
minHIndexinteger–Minimum h-index. Optional. Whole number, 0 or more.
minTopicWorksinteger–Minimum matching papers. Each expert must have at least this many papers on your topic. Whole number, 1 or more papers; default 2.
minWorksinteger–Minimum total works. Optional. Total works on the person's OpenAlex profile. Whole number, 0 or more works.
papersPerExpertinteger–Recent papers per expert. How many of each expert's newest matching papers to list (title, year, DOI, citations). 0 to skip. Whole number, 0 to 10 papers; default 3.
run_max_charge_usdnumber–Optional. Hard spending cap for this call in USD (default from PANDA_MAX_CHARGE_USD, 0.50).
run_max_itemsinteger–Optional. Most results to return from this call (default from PANDA_MAX_ITEMS, 50). Also limits how many results the Actor produces and charges for, unless you set its own max field.
run_timeout_secsinteger–Optional. Seconds to wait for results (default from PANDA_TIMEOUT_SECS, 120; max 300).
sortBystring–Sort. Expert score (default) blends matching papers, their citations and h-index. Or sort by a single number. Choices: expertScore = Expert score; topicWorks = Matching papers; topicCitations = Citat…
topicstringyesResearch topic. What the experts work on, e.g. perovskite solar cells, antimicrobial resistance, large language models in healthcare. Matched against paper titles and abstracts. Use quotes for exact…
yearFrominteger–Papers from year. Optional. Only count matching papers published in or after this year, e.g. 2020 for people active in the field recently. Whole number, 1900 or more.
yearTointeger–Papers to year. Optional. Only count matching papers published up to this year. Whole number, 1900 or more.

No output schema declared.

No examples provided.

topic_trends_tracker ~923

Compare public interest in topics, brands, products or people over time using real Wikipedia pageview counts: daily and monthly views, 30-day and year-over-year change, peaks, spike days, a trend label and a 0-100 interest index. Unlike Google Trends (relative 0-100 only) this reports absolute views; unlike podcast_leads_finder or company_hiring_signals it measures attention, not contacts. Use when you need to size or compare interest, spot spikes, or validate a topic. Do not use for search-keyword volume or social mentions (Wikipedia attention is a proxy), or for topics without a Wikipedia article (returned free as notFound). Parameters: each topic resolves to one article in the chosen language edition; pass a Wikipedia URL to pin an ambiguous name. months sets the window (data from July 2015); year-over-year is always computed. includeDailySeries adds a dailyViews array per row and makes output much larger; leave off unless charting. Returns per item: agentSummary, topic, article, url, interestIndex, trend (surging/rising/flat/falling), totalViews, avgDailyViews, last30dViews, change30dPct, yoyChangePct, peakDate, peakViews, spikes, monthlyViews, optional dailyViews. Behavior: read-only; runs the Apify Actor pandadad0214/topic-trends-tracker (https://apify.com/pandadad0214/topic-trends-tracker) on your own Apify account. Price: $0.002 per topic tracked. Runs that find nothing are not charged per result, and spending is capped per call by run_max_charge_usd (default $0.50). run_max_items (default 50) limits items returned and fills maxTopics when you leave it unset, so it also limits what is billed; setting maxTopics higher can bill more than is returned. Most calls finish in 5-15 s for a few inputs; large inputs or crawls can take minutes, so raise run_timeout_secs (max 300) or shrink the input after a timeout error. Returns JSON {actor, count, items}; long text output is trimmed to whole items, with all items in structuredContent.

NameTypeReqDescription
devicestring–Device. All devices (default), desktop only, or mobile web only. The mobile apps are included under All devices. Choices: all = All devices; desktop = Desktop only; mobile = Mobile web only.
excludeBotsboolean–Count people only (exclude bots and crawlers). On by default. Off counts every request, including automated ones.
includeDailySeriesboolean–Include the daily series in each dataset row. Adds a dailyViews array to every topic row (handy for charts via API). The daily numbers are always in the Excel and DAILY_CSV files either way.
languagestring–Wikipedia language. Language code of the Wikipedia edition to measure: en, de, fr, es, ja, pt, it, nl, pl… Default en. A full Wikipedia link overrides this per topic.
maxTopicsinteger–Maximum topics. You pay per topic tracked, so this also caps the cost. 1 to 1,000; default 100.
monthsinteger–How many months back. Period to analyze, ending yesterday. 1 to 120 months (data starts July 2015); default 12. Year-over-year change is always computed when a year of history exists.
run_max_charge_usdnumber–Optional. Hard spending cap for this call in USD (default from PANDA_MAX_CHARGE_USD, 0.50).
run_max_itemsinteger–Optional. Most results to return from this call (default from PANDA_MAX_ITEMS, 50). Also limits how many results the Actor produces and charges for, unless you set its own max field.
run_timeout_secsinteger–Optional. Seconds to wait for results (default from PANDA_TIMEOUT_SECS, 120; max 300).
topicsarray–Topics. One per line: a topic, brand, product, person, place or a Wikipedia link (https://en.wikipedia.org/wiki/ChatGPT). Each is matched to its Wikipedia article (redirects followed, e.g. Ozempic →…

No output schema declared.

No examples provided.

wayback_snapshot_finder ~974

Find Internet Archive Wayback Machine snapshots for pages or whole domains through the official CDX API: capture timestamps, direct archive URLs, HTTP status, MIME type, content digest, and first/last-seen dates. Unlike website_change_monitor (diffs the live page on a schedule) and reference_checker (verifies scholarly citations), this looks up historical archived copies. Use when you need to cite or recover a dead link, prove what a page said on a date, date when a site or page first appeared, or list a domain's archived URLs. Do not use to fetch or diff live pages (use website_change_monitor) or to download full archived page content. Parameters: urls with matchType exact|domain|prefix, or domains (always domain match); from/to bound the date range (YYYY or YYYYMMDD); closestTo sorts by nearness to one date; collapseDigest (default true) keeps one row per unique content; filterStatus=200 drops redirects and errors; maxResultsPerUrl applies before maxResults, which caps billed snapshots (hard max 5000); includeFirstLast adds extra CDX probes and can add latency. Empty or failed lookups are free; the Archive's CDX API can be slow, so allow up to a minute or more. Returns per item: agentSummary, url, timestamp, timestampReadable, originalUrl, archiveUrl, statusCode, mimeType, digest, length, isFirst, isLast, firstSeen, lastSeen; plus SNAPSHOTS_XLSX/CSV. Behavior: read-only; runs the Apify Actor pandadad0214/wayback-snapshot-finder (https://apify.com/pandadad0214/wayback-snapshot-finder) on your own Apify account. Price: $0.001 per snapshot returned. Runs that find nothing are not charged per result, and spending is capped per call by run_max_charge_usd (default $0.50). run_max_items (default 50) limits items returned and fills maxResults when you leave it unset, so it also limits what is billed; setting maxResults higher can bill more than is returned. Most calls finish in 5-15 s for a few inputs; large inputs or crawls can take minutes, so raise run_timeout_secs (…

NameTypeReqDescription
closestTostring–Closest to timestamp. Optional. Return snapshots nearest this date/time (YYYYMMDD or YYYY-MM-DD). Sorted by closeness.
collapseDigestboolean–Unique content only (collapse by digest). One row per unique page content. On by default.
domainsarray–Domains. Whole domains to search (always uses domain match). Combine with URLs if you want.
filterStatusstring–HTTP status filter. Only keep this status code (e.g. 200). Leave empty for all.
fromstring–From date. Earliest snapshot: YYYY or YYYYMMDD (e.g. 2015 or 20150101).
includeFirstLastboolean–Mark first and last seen. Probe CDX for the earliest and latest capture of each URL and set isFirst / isLast (and firstSeen / lastSeen).
matchTypestring–Match type (for URLs). exact = that URL only; domain = any URL on the host; prefix = that path and below. Choices: exact = Exact URL; domain = Whole domain; prefix = URL prefix.
maxResultsinteger–Maximum results. You pay per snapshot returned, so this also caps the cost. Default 100 (hard max 5000).
maxResultsPerUrlinteger–Max snapshots per URL. Cap per URL/domain before the global max. Default 100.
run_max_charge_usdnumber–Optional. Hard spending cap for this call in USD (default from PANDA_MAX_CHARGE_USD, 0.50).
run_max_itemsinteger–Optional. Most results to return from this call (default from PANDA_MAX_ITEMS, 50). Also limits how many results the Actor produces and charges for, unless you set its own max field.
run_timeout_secsinteger–Optional. Seconds to wait for results (default from PANDA_TIMEOUT_SECS, 120; max 300).
tostring–To date. Latest snapshot: YYYY or YYYYMMDD.
urlsarray–URLs. Pages or hosts to look up (https:// and www. are fine). Duplicates are dropped.

No output schema declared.

No examples provided.

Install

How do I install the Panda Research Tools MCP server?

Panda Research Tools runs locally as a PyPI package, launched with uvx panda-research-tools. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

pypi · panda-research-tools

# add to Claude Code
claude mcp add pandadad0214-panda-research-tools -- uvx panda-research-tools
Common questions

What is the Panda Research Tools MCP server?

Panda Research Tools is an MCP server listed in the public MCP registry as io.github.pandadad0214/panda-research-tools. Scholarly integrity, citations, literature/PubMed, experts, PDF tables, Wayback, trends (8 tools). This page covers its PyPI package (panda-research-tools).

Is the Panda Research Tools MCP server safe to use?

Panda Research Tools scores 68 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 21 November 2026. Its build provenance is signed and verified. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

What tools does the Panda Research Tools MCP server expose?

Panda Research Tools exposes 8 tools: reference_checker, citation_formatter, literature_search, pubmed_evidence_table, researcher_expert_finder, and 3 more. Their descriptions and schemas cost roughly 9,066 tokens of context every time the server is loaded.

Is the Panda Research Tools MCP server still maintained?

Panda Research Tools is still listed as active in the MCP registry. We last reached this channel on 21 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

What licence is the Panda Research Tools MCP server under?

Panda Research Tools declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

What did VerifyMCP test on the Panda Research Tools MCP server?

VerifyMCP ran 15 checks on Panda Research Tools, covering Known CVEs, Install scripts, Dependency health, Source repository and 11 more. One item could not be completed, a gap on our side rather than a finding about Panda Research Tools: Malware scan. Endpoint Security and Transport & Reachability do not apply to a PyPI package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.