Research Integrity Python Packages

proofpath

Verify that the sources behind a claim actually say what the claim says — citation existence, retraction and entailment checks in one offline-first CLI.

refaudit

Checks a .bib file against Crossref, DataCite, arXiv, DBLP, and OpenAlex, and flags DOIs that point to the wrong paper.

paper-preflight

Pre-submission integrity gate for LaTeX papers: every reference verified against real scholarly records. No LLM guessing.

docpluck

Text, section, table and statistic extraction from academic papers (PDF, DOCX, HTML) for meta-science

ghostcite

Deterministic CLI (no LLM in the core check) that catches ghost citations — when the author/year you cited doesn't match the DOI's CrossRef record (+ retraction flags).

citevahti

Run unit tests on your manuscript. Check every manuscript claim against its cited evidence. Human-first, AI-second, Zotero-audited, local-first.

manuscript-guard

Make every number in a scientific manuscript traceable to its source: your results, the literature, or a stated convention. Deterministic checks on Markdown source, built to Word. Alpha.

zotero-provenance

Claude Code plugin: records every source your agent cites into a Zotero collection, tagged with project, context, and date

confound-controls

Confound-matched controls for classifier claims: 1:1 matching, bootstrap AUROC, anchored verdicts, and a refusal when the control is vacuous.

aletheia-probe

Tool for automatically detecting predatory journals and conferences using multiple authoritative data sources

aegis-integrity

Free, local-first academic integrity checker and MCP server: citation hallucination detection, AI-text and plagiarism signals, venue guideline checks (IEEE, ACM, Elsevier). Documents never leave your machine.

citegate

Citation integrity as a CI gate: verify BibTeX references against Crossref & OpenAlex, catch hallucinated citations, get alerted when a cited paper is retracted.

phack

How fast can a coding agent p-hack? Measured: on known-zero data, realistic searches manufacture p<.05 on 33-97% of null draws, in ~1 second each. Instrumented specification-search engine + 11 Claude Code skills for econometric designs; every search leaves a verifiable ledger and an honest p-value. Research and teaching use only.

retractguard

The free, open watchdog for retracted & problematic citations — Retraction Watch + OpenAlex + Crossref, reconciled conservatively. MCP server + web checker, evals-backed.

vahtian

Citation-true research synthesis with Zotero, PubMed intake, provenance tracking, and blinded human–AI dual-rating records.

citeguard-cli

Check citations for retractions, corrections, and expressions of concern via Crossref -- Python CLI/library, MCP server, and GitHub Action sharing one empirically-verified detection algorithm.

warrantos

WarrantOS: CI for claims. Every factual claim in an AI-assisted document ships with a source, a [CITE NEEDED] tag, or a logged BLOCK. No warrant, no ship. Plus warrantos slop, a zero-config AI scaffold-residue scanner. MIT, stdlib-only Python.

quant-integrity

Statistical tests for whether a backtest means anything: deflated Sharpe, PBO, purged CV, reality checks, attestation verification.

auditlite

Reproducibility audit trail for research projects: records runs, checks stale outputs, and flags claims downstream of changed sources via multi-hop change propagation. Python + R.

citeverify

Checks whether the references in a bibliography actually exist, against OpenAlex, Crossref, arXiv and five more indexes.

perception-xalpha-lite

Backtest overfitting audit for factor research: probability of backtest overfitting (PBO), deflated Sharpe ratio, point-in-time data, purged walk-forward. Searches published factor libraries and reports what actually survived costs.

sciaudit

Linter de integridade científica para Python e Jupyter: detecta data leakage, falta de reprodutibilidade e armadilhas metodológicas. Zero dependências.

rigorously

Automated research quality assurance. Catches fabricated citations, overclaimed results, irreproducible numbers.

sciveris

Sciveris re-executes a study and checks whether the finding is real — data leakage, batch/site confounds, effect-size inflation, p-hacking. Latin: "you will have known."