# Simulated Adversarial Peer-Review Protocol

## Scope and disclaimer

This archive is a structured stress test of the public-record reconstruction in `../index.html` and the associated narrated deck. Every reviewer is a **fictional, anonymous functional persona** created for analytical coverage. Some prompts draw on public systems and synthetic-biology themes, but no role represents a named scientist's private views, behavior, questions, mental state, committee conduct, or evaluation of Hersh Bhargava.

The process does not claim access to the private dissertation, live presentation, committee deliberations, unpublished data, or defense outcome. The review cutoff is July 17, 2026.

## Shared evidence packet

- `../index.html`
- `../narration.json`
- `../hersh-defense-brief.pptx`
- `../sources/`
- `EVIDENCE_MANIFEST.tsv`, which freezes the source and extracted-text files supplied to the review process by path, byte size, and SHA-256

## Review dimensions

Each reviewer scores 1–10:

1. Thesis coherence
2. Originality and independence
3. Mechanistic rigor
4. Quantitative and statistical rigor
5. Reproducibility
6. Translational claim calibration
7. Oral-defense readiness

Each review must distinguish:

- verified public evidence;
- reasonable inference;
- missing private evidence;
- overclaim or unsupported language;
- decisive questions that could change the verdict.

## Iterations

1. Generator submission: strongest defensible thesis and defense strategy.
2. Independent adversarial reviews from diverse personas.
3. Author rebuttal and concrete artifact revisions.
4. Second-round reviewers assess whether concerns were resolved.
5. Meta-review synthesizes consensus, disagreements, and required conditions.
6. Final audit checks factual calibration and archive completeness.

## Decision labels

- **Strong accept:** public evidence alone clears the simulated standard.
- **Conditional accept:** scientifically strong, but specified private evidence is required.
- **Major revision:** core thesis plausible, but substantial reasoning or evidence gaps remain.
- **Reject as currently evidenced:** available evidence does not establish a defensible doctoral contribution.

No simulated decision is an actual UCSF or committee decision.