--- name: rt-venue-integrity description: Use before submitting to a venue this repository does not cover, one that arrived by email solicitation, or one whose indexing status matters for a tenure or degree requirement. Runs a source-by-source verification protocol — indexing, publisher, editorial board, fees, retraction history — and reports what each primary source says. It never labels a journal; it reports checks. --- # Venue Integrity Check (rt-venue-integrity) The 743 venues in [`shared-resources/journal-selection/venue-index.tsv`](../../../shared-resources/journal-selection/venue-index.tsv) were each curated into this repository from their own published guidance, so a venue being in the index is at least evidence that someone looked. The moment an author steps outside it — and [`journal-match.md`](../../../shared-resources/journal-selection/journal-match.md)'s coverage-honesty rule requires saying so rather than forcing a poor fit — this repository has nothing to offer, and that is exactly the moment the risk appears: a solicitation email, a plausible-looking title, an APC, and a paper that becomes uncitable and cannot be resubmitted elsewhere. This is the check to run before that submission. ## First: is it actually outside the index? ```bash python3 tools/match_venues.py --title "" --top 5 grep -i "" shared-resources/journal-selection/venue-index.tsv ``` Venues appear under several names, and a Chinese journal's English title or an acronym's expansion often looks unfamiliar when the venue is in fact covered in depth. If it is in the index, stop — the pack answers the question better than this checklist does, and the checks below are for venues nobody here has vetted. ## When to trigger - The target venue is **not** in `venue-index.tsv` and not in a breadth bundle. - The invitation arrived unsolicited, by email, praising the author's "esteemed work". - A degree, tenure or funding rule requires a specific index (SSCI / SCIE / Scopus / CSSCI / 北大核心), and the venue's claim to it has not been verified. - The venue's name closely resembles a well-known journal's. - An APC is quoted and the author is deciding whether to pay it. ## The protocol Verify each item **against the source named**, at the time of asking. Record the source and the date. A claim on the journal's own website is not verification of that claim. | Check | Verify against | A finding worth stopping for | |---|---|---| | Indexing claim (SSCI/SCIE/ESCI) | Clarivate **Master Journal List** search | listed nowhere, or listed only in ESCI while the site claims SCIE | | Indexing claim (Scopus) | Elsevier **Scopus source list** (title + ISSN) | absent, or marked discontinued | | Open-access legitimacy | **DOAJ** entry | absent, or removed | | Chinese index claim | the current **CSSCI** / 北大核心 catalogue for the stated period | claimed for a period it was not listed | | Delisting history | the index's own discontinued-titles list | delisted, and when | | ISSN and title | the **ISSN Portal** | ISSN belongs to a different title | | Publisher identity | publisher's site, membership of **COPE** / OASPA / STM | no traceable publisher, or a membership that cannot be confirmed | | Editorial board | spot-check three named editors at their stated institutions | editors who do not list the role, or cannot be found | | Fees | the journal's own author guidelines | fee disclosed only after acceptance | | Peer review | the stated policy, and the promised timeline | review promised in days; "guaranteed" acceptance | | Retractions | **Retraction Watch** database, PubPeer for the field | a pattern, not a single retraction | Five to ten minutes of checking; the failure it prevents is measured in years. ## What this skill will not do **It ships no list of predatory journals and applies no label.** Those lists are contested, go stale, and calling a named journal predatory is a factual claim about a real publisher that this repository is not in a position to make. The output is a table of *checks*, each with a source and a date, and the author draws the conclusion. Report a check as **verified**, **failed**, or **could not verify** — and treat "could not verify" as its own answer, not as a soft pass. ## Output format ``` 【Venue】name · publisher · ISSN 【Checks】 indexing (SSCI) verified / failed / could not verify — source, date indexing (Scopus) … publisher … editorial board … fees … retractions … 【Unresolved】what could not be verified, and what would resolve it 【What this means for the requirement】whether the degree/tenure index rule is met 【Author's decision】stated as a decision for the author, with the risks named ``` ## Hard rules 1. **Name the source and the date for every check.** An unsourced check is not a check. 2. **Never assert that a journal is predatory, fraudulent or fake.** Report what each primary source says or does not say. 3. **"Could not verify" is a finding**, and for an unfamiliar venue asking for money it is a serious one. Do not round it up to "probably fine". 4. **An indexing requirement is checked against the index, not the journal.** A journal's own claim to be indexed is the thing under test. 5. If the venue **is** in `venue-index.tsv`, this skill is not needed — say so and hand back to `rt-journal-match`. ## Anti-patterns - Judging a journal by its APC. Many reputable open-access journals charge; many predatory ones are cheap. - Treating an unfamiliar name as suspect. Small national and specialist journals are legitimate and often the right home for a paper. - Running the check *after* submission. Withdrawal is not always possible, and simultaneous submission is misconduct. --- *The escape hatch for `rt-journal-match`'s coverage-honesty rule: when the right venue is outside the index, this is what to do instead of forcing a fit. Part of [`Research-Toolkit-Skills`](../../README.md).*