Study Init

Incorporate relevant lessons into the README sections (Assumptions, cqed_sim Gap Analysis, Compute & Resource Strategy).

3. Create the Folder Structure

Create the following directory tree under studies/:

studies/<study_name>/
├── README.md
├── IMPROVEMENTS.md
├── scripts/
│   └── reproducibility_notebook.ipynb  ← REQUIRED (created in Step 6)
├── data/
├── figures/
└── report/

3. Generate the README

Use the README template as the starting point. Fill in:

• Study Title — from the user's description
• Problem Class — classify as OPT, REP, DES, ANA, or a combination
• Motivation — why this study matters; include paper references for REP class
• Goals — numbered, concrete, falsifiable goals
• Methods — which cqed_sim modules and functions will be used (consult the API Reference)
• Analytic Preliminary — the first-principles model plus the closed-form or limiting-case reasoning attempted before numerics, including any controlled approximations and why they are valid
• cqed_sim Gap Analysis — what is needed, what exists upstream, and what must be extended or implemented locally
• Assumptions — physical assumptions, parameter ranges, and convergence criteria
• Compute & Resource Strategy — cost estimate, planned acceleration, and expected bottlenecks
• Expected Outcomes — quantitative success criteria where possible
• Validation — initialize the three gate checks as unchecked items
• Status — set to ACTIVE

If the study consolidates or extends earlier studies, add a ## Study Composition section that maps inherited components to their original source and role.

4. Confirm with the User

Display the generated README and ask whether the scope, goals, and methods look correct before proceeding to implementation.

Rules

• Every study must have a README with all mandatory sections before any code is written.
• The study name must be lowercase with underscores — no spaces, hyphens, or capitals.
• Do not create placeholder scripts or empty Python files. Scripts are written in Step 3 (Implement).
• The README must anchor the planning work: first-principles analytic reasoning, controlled approximations, cqed_sim gap analysis, assumptions, compute strategy, and validation status all belong there, not only in chat.
• Prefer a first-principles Hamiltonian or governing-equation view before simplifying the problem. Any approximation used to make the study tractable must be stated explicitly with its validity condition.
• Every study must include scripts/reproducibility_notebook.ipynb before it can be marked COMPLETE. The notebook is created in Step 6 (see AGENTS.md).
• The reproducibility notebook must expose all tunable parameters (Hilbert space dims, optimizer settings, noise model, cost weights, probe states, sweep ranges) in a single early cell so users can modify and re-run the workflow without rewriting code. See AGENTS.md Step 6 for the full specification.