theoretical-biophysics-modeling-skills
Reusable skills for scientific computing, model documentation, figures, clarification, and agent guidance.
Open source Open in the app JSON README (API)
About
Reusable skills for scientific computing, model documentation, figures, clarification, and agent guidance.
Details
- Kind
- Plugins
- Topic
- Government & public data
- Publisher
- surajinacademia
- Origin
- gemini
- Category
- ferramentas
- Version
- 0.4.0
- Last push
- 2026-09-04T23:34:37Z
- Repository state
- ativo
- Language
- Python
- License
- MIT
- Added
- 2026-09-01 01:00:23
- Updated
- 2026-09-05 02:00:49
- Origin id
surajinacademia/theoretical-biophysics-modeling-skills
README
# Theoretical Biophysics Modeling Skills Five reusable, platform-neutral agent skills. ## Install - **Codex:** ask `$skill-installer` to install the desired folder under `skills/` from this repository. - **Claude Code:** `git clone https://github.com/surajinacademia/theoretical-biophysics-modeling-skills.git ~/.claude/skills/theoretical-biophysics-modeling-skills` - **Gemini CLI:** `gemini extensions install https://github.com/surajinacademia/theoretical-biophysics-modeling-skills` - **Perplexity Computer:** upload a skill as a ZIP with `SKILL.md` at the ZIP root from **Skills → Create skill → Upload a skill**. ## `computational-modeling` Core workflow for implementing, modifying, debugging, validating, or reviewing scientific computations, simulations, data analyses, and figure-generation code. Preserve the stated scientific model, modify the smallest existing pathway, validate the exact scientific claim, and stop. ## `model-documentation` Plan, write, revise, or audit one physics-focused model document from code, approved goals, or an existing methods record. It requires approval before writing and follows a fixed nine-section structure: physical rationale and defined notation first, governing equations next, then reproducible numerical methods and an implementation record. It has no dependency on a writing skill. The skill includes a [fixed template](skills/model-documentation/references/model-document-template.md), a [worked oscillator example](skills/model-documentation/examples/overdamped-harmonic-oscillator/model.md), and a read-only format checker. Independent verification and scientific review remain necessary: a format pass does not establish physical validity. ## `schematic-designer` Use it to create or reconstruct a scientific schematic. It inventories the scientific content, selects Matplotlib, TikZ, or a hybrid workflow, renders the figure, and checks the final PDF and outlined-text SVG. It includes two complete scientific examples, a TikZ starter template, and a 17-entry TikZ gallery with editable sources and rendered specimens. ## `lets-be-clear` Use it to confirm a request before work starts. It restates the goal and scope, asks for confirmation, and does not start the underlying task until the user confirms the unchanged meaning. ## `gotcha` Use it to save explicit agent-performance feedback. It finds the applicable `AGENTS.md`, adds or revises one durable rule, checks authority and security boundaries, and reports the exact change. Each complete skill is under `skills/`. Platform manifests are included at the repository root. The main repository uses the MIT license. The TikZ gallery has entry-specific terms. See `THIRD_PARTY_NOTICES.md` for all third-party material. ## Verify Model Documentation Run the regression tests and adversarial format cases from the repository root: ```bash python3 -B -m unittest discover -s tests -v python3 -B evals/model-documentation/checker_eval.py ``` These check format enforcement, checker error handling, and the oscillator's numerical behavior. They do not measure an agent's workflow reliability or establish experimental validity. The evaluator reports incorrect physical claims as manual-review controls, not as something the format checker can reject.