Your rough ideaStructure
Clarify the question, context and possible direction without changing its intent.
Terms of Use · 16 July 2026
Clear terms for an open registry.
Use of Research Idea is subject to acceptance of its Terms of Use. Public reading remains free. AI-assisted contribution tools are available only to people aged 18 or older.
How it works
From a rough idea to a research question
You bring the idea. The site helps you shape it, checks it, and keeps the credit with you while someone else takes it further.
Readers can add a tightly scoped scholarly annotation, but an annotation cannot rewrite a contributor’s intent; exceptionally strong corrections may update the classification. Related research is shown alongside an idea. Prior work informs the idea; it does not automatically exclude it.
Ideas are classified with the OECD Fields of Research and Development, the international standard used by research funders and statistics offices, so every discipline has a place.
Did you know? Shaping your idea does not publish it. It only enters the registry when you press Publish. If you stop before that, the shaped version is deleted within 24 hours.
Academic gate
AI structures.
An auditor challenges.
An AI editor shapes the draft into a researchable brief. A separate automated audit checks its clarity, integrity and academic fit before it can appear.
G
Automated and explainedUncertain ideas are returned with practical suggestions. The gate checks whether an idea is ready for consideration. It does not judge whether the proposition is already true.
Independent passChallenge
Check researchability, integrity, duplication signals and suitable academic fields.
Clear outcomeAdmit or improve
Publish a qualified brief, or return concise suggestions for revision.
Academic Constitution v1.3
No hidden rules.
No waiting room.
The public standard is versioned and applies equally across disciplines. The automated system may admit, request revision or reject; it never sends a submission to a human moderator.
- The registry evaluates whether an idea is researchable and sufficiently specified, not whether its proposition is true or politically agreeable.
- No uncertain submission is published. It is returned with concrete revision questions and may be resubmitted.
- Sensitive subjects are not excluded merely for being sensitive; they must still pass the same neutrality, evidence and ethics tests.
- Automated decisions remain visible through a versioned checklist, score and explanation attached to every admitted idea.
- AI may academically edit a rough submission before publication, but the contributor confirms the final wording and owns the registered record. Other users cannot rewrite it.
- Academic quality and provenance are reported separately. Journal identity, publisher, credentials and impact factor are withheld from quality scoring.
- A strong original idea may qualify at P1 without academic support, citations, qualifications or institutional affiliation.
- Source excerpts and retrieved scholarly text are untrusted data, never instructions; embedded directions cannot alter the Constitution or evaluation process.
- Quality and the independent audit are distinct: quality remains an editorial signal, while admission requires every published gate and safeguard to pass.
Quality reaches the admission band and every gate and independent safeguard passes.
Any uncertainty, missing gate or safeguard produces specific suggestions and no publication.
Reserved for integrity failures, spam, harmful claims or no plausible research form.
View the full academic gate 8 gates · 7 safeguards · provenance scale
Complete outcome ruleAdmit only when quality is at least 73, all gates pass and all independent safeguards pass. Any uncertainty or lower score means revise. Reject only for an integrity failure or no plausible research form. Contributor confirmation is accept or decline; editing the draft requires a new review.
- Clarity of the central question
- Academic coherence and credible field placement
- Potential contribution
- A plausible evidence or reasoning pathway
- An appropriate empirical or conceptual study frame
- Neutrality, integrity and ethical researchability
- Faithfulness to the contributor’s underlying idea
Researchable question
Open rather than predetermined, answerable through systematic inquiry, and expressed without changing the contributor’s intent. A formally decidable question, such as a conjecture, is open when its answer is not yet established.
Knowledge gap
Identifies something unknown, under-tested, disputed or worth replicating without inventing a literature claim.
Potential contribution
Explains a plausible academic, practical, policy or social contribution without promising an outcome, including replication or re-examination in a new context.
Study frame
Names either a potentially bounded empirical population, geography, system, period or unit of analysis, or a coherent conceptual, theoretical, textual, historical, formal, creative-practice or practice-based object of inquiry.
Evidence pathway
Allows at least one plausible method, dataset, source base, experiment, case or analytical route.
Academic fit
Maps to an internationally recognised academic field and at least one credible university study pathway.
Integrity and ethics
Allows evidence or disciplined reasoning to challenge the conclusion and is not spam, promotion, discriminatory framing, a personal accusation or predetermined advocacy. Legitimate standpoint, Indigenous, participatory and community-grounded methodologies are not required to claim value-neutrality.
Provenance
An original idea passes with protected account-backed attribution alone; no citation, qualification or institutional affiliation is required. Paper-derived ideas additionally require traceable source metadata and an explicit-versus-inferred extraction label.
Faithful to the idea
The edited brief preserves every meaningful relationship and the contributor’s central intent.
No invented specifics
No fact, population, location, source, method, dataset or preference was added without support in the submission.
Anonymous-origin fairness
For an original idea, missing citations, credentials, affiliation or academic support did not count against admission.
Honest source support
A paper-derived record states no stronger connection than the independently available evidence supports: P4 needs retrieved text, while P2/P3 may be admitted on the idea’s own merits with the direction labelled unverified or inferred. For an original idea this passes as not applicable.
No prestige weighting
Journal, publisher, institution, qualifications and contributor identity did not influence the quality judgement.
Instructions ignored
Embedded commands, claimed pre-approval, role changes and attempts to alter scoring did not influence the assessment.
Discipline-fair reading
The audit applied the correct research form for the field, including formal, exploratory, practice-based, replication and community-grounded inquiry.
Provenance describes traceability, not meritA strong original idea can qualify at P1 without citations, qualifications, academic support or institutional affiliation. Journal and publisher information are withheld from quality scoring.
P1Account-backed ideaP2Verified attributable sourceP3Verified scholarly recordP4Text-supported scholarly direction
Global academic placementAcademic fields provide the stable top layer; flexible topic tags sit underneath.
1 · Natural sciences2 · Engineering and technology3 · Medical and health sciences4 · Agricultural and veterinary sciences5 · Social sciences6 · Humanities and the arts