Add AI-generated model summaries and JSON index; refine explain-rulehub-model skill

[Psuedonerd]

Motivation

• Provide machine-readable and human-readable summaries for RuleHub models to support site integration and browsing.
• Create concise, biologist-facing Markdown explanations for many published and tutorial models so users can quickly understand model intent and key behaviors.
• Tighten and clarify the explain-rulehub-model skill to produce consistent Markdown and JSON outputs suitable for automated indexing.

Description

• Add a consolidated JSON index data/ai_summaries.json that lists model entries with model_id, title, bngl_path, yaml_path, markdown_path, summary, and key_events for each model.
• Add many AI-generated Markdown summaries under data/summaries/ and data/tutorial_summaries/, each following the standardized explanatory structure (One-sentence summary, What the model shows, Biological story, Main biological players, Mechanism in plain English, Key modeled events, What the model measures, Expected behavior in plots, Caveats).
• Update the skill documentation skills/explain-rulehub-model/SKILL.md to reorganize inputs, required output structure, strict prohibitions (e.g., no raw BNGL in prose), the JSON schema example, and workflow/quality checklists to ensure consistent summary generation.
• All changes are additive (content and documentation); no BNGL model code was altered.

Testing

• No automated tests were executed as part of this change because the patch primarily adds summary content and documentation rather than executable code or behavior-driving logic.
• The changes are additive files and updated guidance; downstream site integration should validate data/ai_summaries.json and the presence of referenced Markdown files during deployment.

---

Codex Task

[chatgpt-codex-connector]

💡 Codex Review

RuleHub/data/tutorial_summaries/MotivatingexamplecBNGL_aigenerated.md

Line 1 in 3ecbb6e

# Model Explanation: Motivating_example_cmodel

Use the cBNGL model name in cBNGL summaries

This summary is indexed for Tutorials/MotivatingexamplecBNGL/Motivating_example_cBNGL.bngl and that metadata names the model Motivating_example_cBNGL, but the rendered Markdown identifies it as Motivating_example_cmodel throughout. Any consumer following data/ai_summaries.json will show the cBNGL entry under the wrong/nonexistent model name, so the cBNGL summaries need to be regenerated from the metadata title rather than the cmodel variant.

---

RuleHub/data/summaries/Barua2007_aigenerated.md

Line 31 in 3ecbb6e

The readouts distinguish free receptor, receptor bound to SHP2, multiply engaged receptor-SHP2 complexes, open SHP2, and phosphorylated receptor. They show how receptor scaffolding and SHP2 conformational opening control the amount of active phosphatase complex.

Limit Barua 2007 readouts to the actual observable

For this model the BNGL defines only one observable, pYR (Published/Barua2007/Barua_2007.bngl:150-152), so the generated summary is misleading when it says the plots distinguish free receptor, SHP2-bound receptor, multiply engaged complexes, and open SHP2. Users relying on this page would look for outputs the model never records; this section should describe only phosphorylated receptor unless additional observables are added.

---

RuleHub/data/summaries/Dembo1978_aigenerated.md

Line 31 in 3ecbb6e

The readouts track free receptors, ligand-bound receptors, crosslinked polymers, and the fraction of receptors in polymers. The model shows the bell-shaped crosslinking response expected for bivalent ligand-receptor systems.

Do not claim Dembo 1978 outputs polymer fractions

The Dembo 1978 BNGL observables are only free hapten, free receptor, total bonds, free hapten sites, and free receptor sites (Published/Dembo1978/blbr_dembo1978.bngl:126-137); there is no readout for crosslinked polymers or the fraction of receptors in polymers. As written, the summary tells users that the model directly plots polymerization quantities that are not emitted by the committed model, so this should be reworded or the missing measurements added.

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".