Open-source research tools

For HRD & adjacent-field editors

peer-reviewer-finder

Find well-matched, conflict-free, and diverse peer reviewers for a manuscript in human resource development and adjacent fields.

Python · JavaScript · OpenAlex · MIT

Given a manuscript's title, abstract, and a few keywords, along with the submitting authors' institutions, it searches a registry of 97 journals spanning human resource development and six adjacent disciplines — adult and continuing education, management and organizational behavior, industrial-organizational psychology, higher education, career and workforce development, and international and comparative education — on OpenAlex for scholars whose published work genuinely matches. It screens out conflicts of interest, then returns a relevance-ranked panel with institutional and national diversity built in. The editor still decides; the tool supplies the evidence and a defensible shortlist.

• ▪Searches 97 journals across HRD and six adjacent disciplines via OpenAlex
• ▪Screens conflicts of interest by shared institution and co-authorship
• ▪Proposes a diverse, relevance-ranked shortlist with alternates
• ▪Runs locally or in your browser — sends no manuscript text to any AI

For authors, reviewers, and editors

claim-verifier

Check whether the sources you cite actually say what you claim they say.

Python · Ollama · MIT

It reads a manuscript, finds every sentence that carries a citation, and pulls the passages from that source's full text that bear on the sentence. A language model then compares the two and returns a graded assessment, from consistent to possibly conflicting, with the passage it relied on quoted underneath. The report is ranked so the citations worth your attention come first. Quotation-error studies in medicine put source misstatement at roughly one in six citations, and few of us check, because checking one citation by hand means finding the source and reading enough of it to be sure. This makes that check cheap. It renders no verdicts: a claim marked as not found means only that support was not found in the passages retrieved, and you are sent to the source to judge.

• ▪Extracts every citation-bearing sentence and matches it to the cited work's full text
• ▪Grades each pairing from consistent to possibly conflicting, quoting the passage behind the call
• ▪Runs on a local model by default, so a manuscript under review never leaves your machine
• ▪Flags rather than judges: what it cannot confirm is a citation to check, not one to fix

claimverify — synthetic example, local model

### zarestky-2024
> Zarestky (2024) found that women's participation in employer
  training lagged men's by ten percentage points.

Verdict: partially consistent (check the shift)
The manuscript states a ten percentage point gap, while the source
reports a nine percentage point gap.

Evidence from the source: "We found that women's participation in
employer-provided training lagged men's by nine percentage points,
and the gap was concentrated in manufacturing."

For scholars who write with AI

Scholarly Voice and Identity Interviewer

Interview yourself once and get two reference documents that give an AI assistant your scholarly voice and intellectual perspectives, in place of the flat, generic default.

Prompt · Markdown · CC BY 4.0

Paste it into Claude and it reads your CV and a spread of your published work, reports what your writing already reveals, then interviews you where the corpus leaves gaps. It produces a Researcher Identity Profile (your animating questions, distinctive position, theoretical toolkit, and quality filters) and a Scholarly Voice Profile (your sentence mechanics, a genre-by-genre reading of your prose, and a register guide). Keep both and paste them into any later chat to draft, revise, or copy-edit as yourself. It is the academic descendant of Ruben Hassid's Taste Interviewer prompt.

• ▪Reads your real published work across genres before it asks a single question
• ▪Captures the intellectual perspectives under your prose, the questions and positions that make it yours
• ▪Calibrates your voice from journal article to keynote to grant
• ▪Records the common AI writing practices you want to avoid, as a check on later drafts