Universal Evidence-Bound Peer-Review Support Tool for Universities, Researchers, and Hobbyists Using Free ChatGPT, Claude, and Gemini

This Zenodo record establishes an independent, open-methods pilot for reviewing dissertations and other research claims through the V6A Falsifiability Sheet. The method is designed for university students, academic researchers, independent researchers, hobbyists, reviewers, and anyone interested in testing a claim through a transparent evidence trail. Its first planned live dissertation case involves a postgraduate student affiliated with Alexandria University, Egypt. The participant’s identity and case details will be added through a later update only after the case is completed and public attribution is authorized. The workflow examines one bounded claim at a time through three separate PDFs. The first contains the completed human R1 Falsifiability Sheet and an illustrated teaching comic explaining how the system works. The second contains a Quick Teaching Guide and the isolated Evidence Packet, which defines the only evidence the AI may use for R2. The third is the blank V6A University Output Sheet, which the AI completes and returns as a finished PDF. Participants upload the three files separately to the free version of ChatGPT, Claude, or Gemini and send one short prompt assigning each file its role. The AI preserves the human R1 record, ignores the R1 conclusion while independently deciding R2, reviews only the supplied Evidence Packet, applies the printed comparison rules, and creates a completed PDF preserving the exact claim, evidence trail, R1 review, R2 review, disagreement, uncertainty, and final Green, Yellow, Red, or STOP result. The isolated Evidence Packet is the method’s central AI-safety control. The AI must not browse outside the packet, introduce outside sources, fill gaps from model memory, invent citations or evidence, or treat its own output as evidence. When essential evidence is missing, unreadable, or insufficient, the required outcome is STOP, not a guessed answer. The method therefore treats AI as a constrained second reviewer rather than as an authority. V6A does not replace supervisors, specialists, formal peer reviewers, ethical review, or university requirements. Its purpose is to make the review path visible: what claim was tested, what evidence was allowed, how the human reviewer judged it, how the AI reviewer judged it, and how the final result was reached. The Falsifiability Sheet was designed to turn a compressed research conclusion into a bounded, evidence-linked, reviewable, and revisable claim record. An earlier British Museum EA1006 control test showed that the same completed R1 record, bounded Evidence Packet, blank output sheet, and short prompt could operate across Gemini, ChatGPT, and Claude. All three systems recognized the principal file roles and returned completed PDF records, supporting cross-provider portability at proof-of-concept level. That test was not full validation. It used one known Green case, R1 remained visible during R2, each provider was tested once, Yellow, Red, and STOP conditions were not yet tested, and differences appeared among the outputs. These limitations remain visible in the research record rather than being hidden. The planned Alexandria University dissertation case will extend the method from a controlled museum baseline into a live postgraduate research setting. This is an independent pilot and does not claim official endorsement, partnership, institutional adoption, approved dissertation status, or formal peer-review recognition by Alexandria University or any other institution. Exact Message Prompt: "Using only the uploads, treat the photo as the completed R1 record, ignore the R1 outcome when deciding R2, use only the Evidence Packet for R2, complete the University Output Sheet, and create the completed PDF." Additional Description — Notes Pilot Status and Participation This record establishes the universal pilot framework and will be updated as live cases, replications, revisions, and completed outputs become available. The first planned university dissertation case involves a postgraduate student affiliated with Alexandria University, Egypt. The Alexandria pilot has not yet been presented as a completed case within this record. The participant’s identification and case-specific details will remain withheld during the initial test. They will be added through a future Zenodo update following completion and only when public attribution has been authorized. Additional Alexandria University students, students at other universities, academic researchers, independent researchers, specialists, reviewers, hobbyists, and interested members of the public may independently test the workflow using their own bounded claims and evidence. Participation does not establish: A formal university partnership An approved dissertation procedure Institutional adoption Ethical approval Academic accreditation Official peer-review status Endorsement by Alexandria University or another institution Each participant remains responsible for following the academic, legal, ethical, copyright, privacy, and research requirements that apply to their own project, institution, evidence, and location. The Zenodo DOI is the permanent archive for the pilot. Updated versions may include revised forms, corrected instructions, replication reports, completed case studies, blank templates, evidence packets, documented limitations, and other materials related to the method. The Echoes of the Script: Open Lab LinkedIn group serves as an optional discussion forum for questions, replication reports, reviewer exchange, voluntary feedback, and public updates. Participation in the LinkedIn forum is not required to use or replicate the method.