Skip to content
Apart Research
Apart Research
Contact us
Sprint projectMay 24, 2026Vadodara

Veridict

Sumit Vekariya · Team ZKred

Submitted to The Secure Program Synthesis Hackathon. Sprint projects are early-stage work by participants, not Apart Research publications.

Read the report

Report: Veridict

Open in a new tab (opens in new tab)
Recording (opens in new tab)Code (opens in new tab)
Share

Veridict is a merge gate for AI-synthesized code that decouples qualified-reviewer-ness from identity. A natural-language spec is sent to Claude, which generates an implementation, a pytest suite, and a Z3 invariant file. The issuer mints a Longfellow ZK credential only after all three formal layers pass, and N anonymous proofs from credentialed reviewers flip the GitHub branch-protection gate that authorizes the merge. Built end-to-end on a live public repo with active branch protection; same Google ZK primitive used for anonymous age verification on mobile driver's licences, repurposed for code review.

Reviews

Judging this Sprint?

Review this project

Your public critique appears on this page without your name. Your private critique is not published; only the Apart team reads it. If you agree below, we share your review with grantmaking.ai (opens in new tab) and the Transformative AI Fund so strong projects can be funded.

Not shown on this page.

Shown on this page, without your name.

Only the Apart team reads this, and funders if you agree below.

Share my name publicly on grantmaking.ai *
Share my private critique with funders *

How much would this matter for AI safety if it worked? How innovative is it? For scores of 4-5: is this actually new to the field, or replicating recent work?

Scoring guide
  1. 1Negligible. No clear problem addressed, or no meaningful novelty.
  2. 2Limited. Addresses a real problem but with a generic or well-trodden approach. Incremental at best.
  3. 3Moderate. Clear problem with a reasonable approach; some novelty in framing or method beyond routine application of existing tools.
  4. 4Significant. Important problem with an original approach, or identifies a neglected problem area. A valuable contribution others could build on.
  5. 5Exceptional. Tackles a critical AI safety problem with a genuinely novel approach, or opens a new research direction. Clear theory of change. You'd be excited to share this with researchers in the area.

How sound are methodology, implementation, and findings?

Scoring guide
  1. 1Seriously flawed. Methodology broken, results uninterpretable, or implementation doesn't work.
  2. 2Weak. Approach has significant gaps: missing validation, flawed experimental design, or incomplete implementation.
  3. 3Competent. Technically solid given the short duration. Methodology makes sense, results are interpretable, limitations acknowledged, work builds toward clear conclusions.
  4. 4Strong. Thorough methodology with convincing validation. Results clearly support conclusions. Immediately useful for future work.
  5. 5Exceptional. Ambitious scope executed rigorously. Surprising findings, novel methods, or unusually robust validation.

How clearly are work, findings, and impact potential communicated?

Scoring guide
  1. 1Incomprehensible. Cannot determine what the project is actually claiming or doing.
  2. 2Hard to follow. Key information buried, missing, or diluted by excessive length. Significant effort to extract main points.
  3. 3Clear enough. Can understand the problem, approach, and results without undue effort. Core content clearly present: problem, method, findings, limitations.
  4. 4Well presented. Easy to follow, well-structured, appropriate level of detail. Target audience would get it quickly.
  5. 5Exceptionally clear. A pleasure to read. Complex ideas made accessible. Could serve as a model for how to present this type of work.

  1. While the GitHub integration and overall user experience are well executed, the core AI security design needs more work to be done. The current Z3 check only validates that the specification is internally consistent; it does not prove that the generated Python code actually follows that specification. This leaves room for incorrect or even malicious code to pass review. In addition, the trust model around zero-knowledge proofs is weakened by server-side key generation and the absence of nullifiers, which could allow duplicate or spoofed approvals. Strengthening the system would require true code-to-spec verification and moving all ZK proof generation fully to the client side.

  2. This is a cool project and very timely. My main feedback would be that this could be better integrated with Git (as opposed to GitHub) and existing infrastructure like Sign-offs. Anyways, this is good work.

Cite this project

@misc{vekariya2026veridict,
  title = {{Veridict}},
  author = {Sumit Vekariya},
  year = {2026},
  month = may,
  note = {Submitted to The Secure Program Synthesis Hackathon, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/veridict-e67q}},
  url = {https://apartresearch.com/sprints/projects/veridict-e67q}
}

More from The Secure Program Synthesis Hackathon

All 61 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026