{
  "markdown": "# Theoretical Biophysics Modeling Skills\n\nFive reusable, platform-neutral agent skills.\n\n## Install\n\n- **Codex:** ask `$skill-installer` to install the desired folder under `skills/` from this repository.\n- **Claude Code:** `git clone https://github.com/surajinacademia/theoretical-biophysics-modeling-skills.git ~/.claude/skills/theoretical-biophysics-modeling-skills`\n- **Gemini CLI:** `gemini extensions install https://github.com/surajinacademia/theoretical-biophysics-modeling-skills`\n- **Perplexity Computer:** upload a skill as a ZIP with `SKILL.md` at the ZIP root from **Skills → Create skill → Upload a skill**.\n\n## `computational-modeling`\n\nCore workflow for implementing, modifying, debugging, validating, or\nreviewing scientific computations, simulations, data analyses, and\nfigure-generation code.\n\nPreserve the stated scientific model, modify the smallest existing pathway,\nvalidate the exact scientific claim, and stop.\n\n## `model-documentation`\n\nPlan, write, revise, or audit one physics-focused model document from code,\napproved goals, or an existing methods record. It requires approval before\nwriting and follows a fixed nine-section structure: physical rationale and\ndefined notation first, governing equations next, then reproducible numerical\nmethods and an implementation record. It has no dependency on a writing skill.\n\nThe skill includes a [fixed template](skills/model-documentation/references/model-document-template.md),\na [worked oscillator example](skills/model-documentation/examples/overdamped-harmonic-oscillator/model.md),\nand a read-only format checker. Independent verification and scientific review\nremain necessary: a format pass does not establish physical validity.\n\n## `schematic-designer`\n\nUse it to create or reconstruct a scientific schematic. It inventories the\nscientific content, selects Matplotlib, TikZ, or a hybrid workflow, renders the\nfigure, and checks the final PDF and outlined-text SVG. It includes two complete\nscientific examples, a TikZ starter template, and a 17-entry TikZ gallery with\neditable sources and rendered specimens.\n\n## `lets-be-clear`\n\nUse it to confirm a request before work starts. It restates the goal and scope,\nasks for confirmation, and does not start the underlying task until the user\nconfirms the unchanged meaning.\n\n## `gotcha`\n\nUse it to save explicit agent-performance feedback. It finds the applicable\n`AGENTS.md`, adds or revises one durable rule, checks authority and security\nboundaries, and reports the exact change.\n\nEach complete skill is under `skills/`. Platform manifests are included at the\nrepository root.\n\nThe main repository uses the MIT license. The TikZ gallery has entry-specific\nterms. See `THIRD_PARTY_NOTICES.md` for all third-party material.\n\n## Verify Model Documentation\n\nRun the regression tests and adversarial format cases from the repository root:\n\n```bash\npython3 -B -m unittest discover -s tests -v\npython3 -B evals/model-documentation/checker_eval.py\n```\n\nThese check format enforcement, checker error handling, and the oscillator's\nnumerical behavior. They do not measure an agent's workflow reliability or\nestablish experimental validity. The evaluator reports incorrect physical claims\nas manual-review controls, not as something the format checker can reject.\n",
  "bytes": 3283,
  "sha": "6891c380604631034320f8786214e07ec197612585c693940942862072a10d06",
  "repo_slug": "surajinacademia/theoretical-biophysics-modeling-skills",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_surajinacademia_theoretical_biophysics_m_91cff38f/readme"
}