[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7APT75CSADVRCQ4ZRVTGAT4":3,"skill-same:01M7APT75CSADVRCQ4ZRVTGAT4":48},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M7APT75CSADVRCQ4ZRVTGAT4","drugdev-qsar-admet-generative-models-and-validation",null,"Use for the machine learning layer: QSAR and property prediction, ADMET prediction and its applicability domain, generative models for molecule design and what they actually produce, active learning and the design-make-test-analyse loop, the benchmarks and why leaderboard performance transfers so poorly, and validation that means something — prospective tests, baselines and the checks that catch a leaking model.","the-vibey-project\u002Fvibey","the-vibey-project","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F329298758?v=4","01M7AJHGZ9722CTJ4JNQZVRH0Q","the vibey project","01M7AJHGZVMG2K7TGXQ6XFHVG3",5,"MIT","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fmedicine-design-software-development\u002Fskills\u002Fdrugdev-qsar-admet-generative-models-and-validation\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: drugdev-qsar-admet-generative-models-and-validation\ndescription: \"Use for the machine learning layer: QSAR and property prediction, ADMET prediction and its applicability domain, generative models for molecule design and what they actually produce, active learning and the design-make-test-analyse loop, the benchmarks and why leaderboard performance transfers so poorly, and validation that means something — prospective tests, baselines and the checks that catch a leaking model.\"\n---\n\n# Drug Discovery Software: QSAR and Property Prediction, ADMET Prediction, Generative Models, Active Learning and the DMTA Loop, Benchmarks, and Validation That Means Something\n\n> **Part 4 of 6** of the *Medicine Design Software Development* reference (plugin `medicine-design-software-development`), covering §13–§18. Sibling skills: `drugdev-pipeline-targets-and-drug-likeness` (§0–§3), `drugdev-representation-cheminformatics-data-quality-and-leakage` (§4–§7), `drugdev-protein-structure-docking-and-molecular-dynamics` (§8–§12), `drugdev-pipeline-engineering-compute-and-regulated-software` (§19–§23), `drugdev-reference` (§24–§29). Section numbers are shared across the set; a reference written as §N → `skill` points into that sibling skill.\n>\n> **Currency:** The chemistry and physics are stable. Two areas moved. See §24 → `drugdev-reference` for the AI drug discovery clinical evidence, and the FDA and EMA regulatory framework for AI.\n\n> **⚠️ Written for software engineers entering computational drug discovery — a domain\n> where the code is easy and the EVALUATION is brutally hard.** ⚠️ **You can build a\n> molecular property predictor in an afternoon and get a beautiful R². It will very\n> probably be measuring leakage, not chemistry** (§7 → `drugdev-representation-cheminformatics-data-quality-and-leakage`).\n>\n> **Complements an AI\u002FML reference (model architectures), a healthcare\u002Fclinical reference\n> (the trial end), and a scientific-computing reference (HPC and reproducibility).**\n>\n> **⚠️ GOTCHA** boxes mark the evaluation traps and the places where a correct-looking\n> result is meaningless.\n>\n> **The three ideas that organize this document:**\n> 1. **⚠️ The bottleneck is BIOLOGY, not compute** (§1 → `drugdev-pipeline-targets-and-drug-likeness`, §24.1 → `drugdev-reference`). **Roughly 90% of candidates\n>    entering clinical trials fail, mostly for efficacy and toxicity reasons that appear\n>    only in humans — and no amount of upstream modelling has yet moved that number.**\n> 2. **⚠️ Random train\u002Ftest splits are almost always WRONG here** (§7 → `drugdev-representation-cheminformatics-data-quality-and-leakage`). **Chemical data is\n>    clustered into analogue series, so random splitting tests interpolation within a\n>    series and reports it as generalization. This is the field's defining methodological\n>    failure.**\n> 3. **⚠️ Prediction is cheap; VALIDATION is the product** (§18, §24.2 → `drugdev-reference`). **In a regulated\n>    context, an unvalidated model is not evidence — the FDA framework is explicitly about\n>    credibility for a stated context of use, not about accuracy in the abstract.**\n\n---\n\n## §13. QSAR and Property Prediction\n\n**⚠️ Predicting activity or a property from structure — the oldest ML application in the\nfield, dating to the 1960s.**\n```\n⚠️ CLASSICAL  descriptors\u002Ffingerprints + random forest, SVM or\n   gradient boosting. ⚠️ STILL HIGHLY COMPETITIVE, and frequently\n   beats deep learning on small datasets — which is most datasets here\n⚠️ DEEP  GNNs (message passing on the molecular graph),\n   transformers on SMILES, ⚠️ pretrained molecular foundation models\n⚠️ MULTITASK  related endpoints trained together often help\n```\n> **⚠️ GOTCHA — ACTIVITY CLIFFS break the core assumption.** ⚠️ **QSAR assumes similar\n> structures have similar activity; activity cliffs are pairs of near-identical molecules\n> with order-of-magnitude activity differences, and they are exactly the cases medicinal\n> chemists care about.** **⚠️ Models systematically fail on them, and aggregate metrics\n> hide this because cliffs are a minority of pairs.**\n> **⚠️ Report performance on cliff pairs separately if the model will be used for lead\n> optimization.**\n\n**⚠️ Honest expectations**: ⚠️ **on well-populated endpoints with good data, models are\nuseful triage tools.** **⚠️ They rarely replace an assay, and their value is in ORDERING\nwhat to test next** (§16), **not in producing numbers to quote.**\n\n---\n\n## §14. ADMET Prediction\n\n```\nABSORPTION  solubility, permeability (Caco-2, PAMPA), efflux (P-gp)\nDISTRIBUTION  plasma protein binding, ⚠️ blood-brain barrier, volume\nMETABOLISM  ⚠️ CYP450 inhibition and substrate; metabolic stability;\n   ⚠️ SITE OF METABOLISM prediction\nEXCRETION  clearance, half-life\n⚠️ TOXICITY  ⚠️ hERG (cardiac liability — the classic killer),\n   hepatotoxicity, Ames mutagenicity, reactive metabolites\n```\n**⚠️ ADMET models are among the most USEFUL in practice** — ⚠️ **because the endpoints are\nmeasured consistently in-house, the datasets are large, and early filtering has genuine\nvalue.** **⚠️ hERG and CYP models in particular are standard.**\n**⚠️ The caution**: ⚠️ **in vitro endpoints are proxies for in vivo behaviour, and in vivo\nis a proxy for human.** **⚠️ Each step loses information, which is why physiologically-based\npharmacokinetic (PBPK) modelling exists to bridge them** — **and PBPK is itself an area\nwhere regulators accept computational evidence** (§24.2 → `drugdev-reference`).\n\n---\n\n## §15. Generative Models\n\n**⚠️ De novo design — proposing new molecules rather than filtering existing ones.**\n```\nAPPROACHES  VAEs · GANs · autoregressive (SMILES\u002FSELFIES) ·\n   ⚠️ reinforcement learning against a scoring function ·\n   ⚠️ diffusion models (increasingly, especially 3D structure-based)\n⚠️ CONDITIONING  on a target pocket, on desired properties,\n   scaffold-constrained, fragment growing\u002Flinking\n```\n> **⚠️ GOTCHA — generation is not the hard part; SCORING is.** ⚠️ **A generative model\n> optimizes whatever objective you give it, and if that objective is a docking score\n> (§10 → `drugdev-protein-structure-docking-and-molecular-dynamics`) or a QSAR model (§13), the generator will find that function's failure modes.**\n> **⚠️ This is reward hacking with molecules — you get compounds that score brilliantly\n> and are inactive, unstable or unmakeable.**\n> **⚠️ Therefore: SYNTHESIZABILITY constraints (SA score, retrosynthesis-aware generation,\n> or generating only from purchasable building blocks) are not a nicety.** ⚠️ **And\n> \"novelty\" metrics are nearly meaningless — it is trivial to generate novel molecules and\n> hard to generate novel USEFUL ones.**\n\n**⚠️ Retrosynthesis prediction and computer-aided synthesis planning** (⚠️ **ASKCOS,\nAiZynthFinder, commercial tools**) **close the loop, and ⚠️ they are the practical gate on\nwhether a designed molecule can actually be made.**\n\n---\n\n## §16. Active Learning and the DMTA Loop\n\n**⚠️ The framing that makes ML in this domain make sense** (§1 → `drugdev-pipeline-targets-and-drug-likeness`): ⚠️ **you are not building a\nmodel to be accurate; you are building a model to CHOOSE WHAT TO TEST NEXT.**\n```\n⚠️ The loop: model → select compounds → SYNTHESIZE AND ASSAY →\n   retrain → repeat\n⚠️ ACQUISITION  exploit (best predicted) vs explore (most uncertain) vs\n   diverse batch selection. ⚠️ Batch selection matters because you make\n   compounds in batches, not one at a time\n⚠️ THEREFORE UNCERTAINTY QUANTIFICATION IS A FIRST-CLASS REQUIREMENT,\n   not a nice-to-have. ⚠️ Ensembles, conformal prediction, Gaussian\n   processes on fingerprints\n⚠️ THE METRIC THAT MATTERS  ⚠️ how many DMTA cycles to reach the target\n   profile — not test-set R²\n```\n**⚠️ Self-driving labs and closed-loop automation** connect this to robotic synthesis and\nassay — ⚠️ **and the real bottleneck becomes assay throughput and reliability, not\nprediction.**\n\n---\n\n## §17. ⚠️ Benchmarks\n\n**⚠️ Common ones**: **MoleculeNet, Therapeutics Data Commons (TDC), DUD-E and LIT-PCBA\n(virtual screening), CASF and PDBbind (scoring), GuacaMol and MOSES (generative), and\npolaris-style curated benchmarks.**\n> **⚠️ GOTCHA — most published benchmark improvements do not transfer, and there are\n> documented structural reasons.** ⚠️ **MoleculeNet's random splits are known to inflate\n> results (§7 → `drugdev-representation-cheminformatics-data-quality-and-leakage`); DUD-E's decoys are separable from actives by trivial properties, so models\n> learn the decoy generation procedure; PDBbind has train-test protein overlap; and\n> generative benchmarks reward distributional metrics that don't correspond to usefulness.**\n> **⚠️ Independent reanalyses have repeatedly found that simple baselines — random forest\n> on ECFP fingerprints — match or beat elaborate architectures once splits are fixed.**\n> **⚠️ Always run that baseline. If your model doesn't beat it, you have learned something\n> important.**\n\n---\n\n## §18. ⚠️ Validation That Means Something\n\n```\n⚠️ RETROSPECTIVE  time split (§7); ⚠️ external test set from a\n   different lab; performance stratified by applicability domain\n⚠️ PROSPECTIVE  ⚠️ THE ONLY REAL TEST — predict, then synthesize and\n   assay, and report the outcome including failures\n⚠️ THE COMPARISON THAT MATTERS  not \"better than random\" but\n   ⚠️ \"better than a medicinal chemist\" or \"better than the cheap baseline\"\n⚠️ CALIBRATION  are the confidence estimates honest? (§16)\n⚠️ ERROR ANALYSIS  where does it fail, and is that a systematic class?\n```\n**⚠️ The publication asymmetry to be aware of**: ⚠️ **prospective validations that failed\nare rarely published, so the visible literature overstates the field.** ⚠️ **This is the\nsame problem as §24.1 → `drugdev-reference` at a smaller scale.**\n\n---\n\n# PART V — ENGINEERING\n","# Drug Discovery Software: QSAR and Property Prediction, ADMET Prediction, Generative Models, Active Learning and the DMTA Loop, Benchmarks, and Validation That Means Something\n\n> **Part 4 of 6** of the *Medicine Design Software Development* reference (plugin `medicine-design-software-development`), covering §13–§18. Sibling skills: `drugdev-pipeline-targets-and-drug-likeness` (§0–§3), `drugdev-representation-cheminformatics-data-quality-and-leakage` (§4–§7), `drugdev-protein-structure-docking-and-molecular-dynamics` (§8–§12), `drugdev-pipeline-engineering-compute-and-regulated-software` (§19–§23), `drugdev-reference` (§24–§29). Section numbers are shared across the set; a reference written as §N → `skill` points into that sibling skill.\n>\n> **Currency:** The chemistry and physics are stable. Two areas moved. See §24 → `drugdev-reference` for the AI drug discovery clinical evidence, and the FDA and EMA regulatory framework for AI.\n\n> **⚠️ Written for software engineers entering computational drug discovery — a domain\n> where the code is easy and the EVALUATION is brutally hard.** ⚠️ **You can build a\n> molecular property predictor in an afternoon and get a beautiful R². It will very\n> probably be measuring leakage, not chemistry** (§7 → `drugdev-representation-cheminformatics-data-quality-and-leakage`).\n>\n> **Complements an AI\u002FML reference (model architectures), a healthcare\u002Fclinical reference\n> (the trial end), and a scientific-computing reference (HPC and reproducibility).**\n>\n> **⚠️ GOTCHA** boxes mark the evaluation traps and the places where a correct-looking\n> result is meaningless.\n>\n> **The three ideas that organize this document:**\n> 1. **⚠️ The bottleneck is BIOLOGY, not compute** (§1 → `drugdev-pipeline-targets-and-drug-likeness`, §24.1 → `drugdev-reference`). **Roughly 90% of candidates\n>    entering clinical trials fail, mostly for efficacy and toxicity reasons that appear\n>    only in humans — and no amount of upstream modelling has yet moved that number.**\n> 2. **⚠️ Random train\u002Ftest splits are almost always WRONG here** (§7 → `drugdev-representation-cheminformatics-data-quality-and-leakage`). **Chemical data is\n>    clustered into analogue series, so random splitting tests interpolation within a\n>    series and reports it as generalization. This is the field's defining methodological\n>    failure.**\n> 3. **⚠️ Prediction is cheap; VALIDATION is the product** (§18, §24.2 → `drugdev-reference`). **In a regulated\n>    context, an unvalidated model is not evidence — the FDA framework is explicitly about\n>    credibility for a stated context of use, not about accuracy in the abstract.**\n\n---\n\n## §13. QSAR and Property Prediction\n\n**⚠️ Predicting activity or a property from structure — the oldest ML application in the\nfield, dating to the 1960s.**\n```\n⚠️ CLASSICAL  descriptors\u002Ffingerprints + random forest, SVM or\n   gradient boosting. ⚠️ STILL HIGHLY COMPETITIVE, and frequently\n   beats deep learning on small datasets — which is most datasets here\n⚠️ DEEP  GNNs (message passing on the molecular graph),\n   transformers on SMILES, ⚠️ pretrained molecular foundation models\n⚠️ MULTITASK  related endpoints trained together often help\n```\n> **⚠️ GOTCHA — ACTIVITY CLIFFS break the core assumption.** ⚠️ **QSAR assumes similar\n> structures have similar activity; activity cliffs are pairs of near-identical molecules\n> with order-of-magnitude activity differences, and they are exactly the cases medicinal\n> chemists care about.** **⚠️ Models systematically fail on them, and aggregate metrics\n> hide this because cliffs are a minority of pairs.**\n> **⚠️ Report performance on cliff pairs separately if the model will be used for lead\n> optimization.**\n\n**⚠️ Honest expectations**: ⚠️ **on well-populated endpoints with good data, models are\nuseful triage tools.** **⚠️ They rarely replace an assay, and their value is in ORDERING\nwhat to test next** (§16), **not in producing numbers to quote.**\n\n---\n\n## §14. ADMET Prediction\n\n```\nABSORPTION  solubility, permeability (Caco-2, PAMPA), efflux (P-gp)\nDISTRIBUTION  plasma protein binding, ⚠️ blood-brain barrier, volume\nMETABOLISM  ⚠️ CYP450 inhibition and substrate; metabolic stability;\n   ⚠️ SITE OF METABOLISM prediction\nEXCRETION  clearance, half-life\n⚠️ TOXICITY  ⚠️ hERG (cardiac liability — the classic killer),\n   hepatotoxicity, Ames mutagenicity, reactive metabolites\n```\n**⚠️ ADMET models are among the most USEFUL in practice** — ⚠️ **because the endpoints are\nmeasured consistently in-house, the datasets are large, and early filtering has genuine\nvalue.** **⚠️ hERG and CYP models in particular are standard.**\n**⚠️ The caution**: ⚠️ **in vitro endpoints are proxies for in vivo behaviour, and in vivo\nis a proxy for human.** **⚠️ Each step loses information, which is why physiologically-based\npharmacokinetic (PBPK) modelling exists to bridge them** — **and PBPK is itself an area\nwhere regulators accept computational evidence** (§24.2 → `drugdev-reference`).\n\n---\n\n## §15. Generative Models\n\n**⚠️ De novo design — proposing new molecules rather than filtering existing ones.**\n```\nAPPROACHES  VAEs · GANs · autoregressive (SMILES\u002FSELFIES) ·\n   ⚠️ reinforcement learning against a scoring function ·\n   ⚠️ diffusion models (increasingly, especially 3D structure-based)\n⚠️ CONDITIONING  on a target pocket, on desired properties,\n   scaffold-constrained, fragment growing\u002Flinking\n```\n> **⚠️ GOTCHA — generation is not the hard part; SCORING is.** ⚠️ **A generative model\n> optimizes whatever objective you give it, and if that objective is a docking score\n> (§10 → `drugdev-protein-structure-docking-and-molecular-dynamics`) or a QSAR model (§13), the generator will find that function's failure modes.**\n> **⚠️ This is reward hacking with molecules — you get compounds that score brilliantly\n> and are inactive, unstable or unmakeable.**\n> **⚠️ Therefore: SYNTHESIZABILITY constraints (SA score, retrosynthesis-aware generation,\n> or generating only from purchasable building blocks) are not a nicety.** ⚠️ **And\n> \"novelty\" metrics are nearly meaningless — it is trivial to generate novel molecules and\n> hard to generate novel USEFUL ones.**\n\n**⚠️ Retrosynthesis prediction and computer-aided synthesis planning** (⚠️ **ASKCOS,\nAiZynthFinder, commercial tools**) **close the loop, and ⚠️ they are the practical gate on\nwhether a designed molecule can actually be made.**\n\n---\n\n## §16. Active Learning and the DMTA Loop\n\n**⚠️ The framing that makes ML in this domain make sense** (§1 → `drugdev-pipeline-targets-and-drug-likeness`): ⚠️ **you are not building a\nmodel to be accurate; you are building a model to CHOOSE WHAT TO TEST NEXT.**\n```\n⚠️ The loop: model → select compounds → SYNTHESIZE AND ASSAY →\n   retrain → repeat\n⚠️ ACQUISITION  exploit (best predicted) vs explore (most uncertain) vs\n   diverse batch selection. ⚠️ Batch selection matters because you make\n   compounds in batches, not one at a time\n⚠️ THEREFORE UNCERTAINTY QUANTIFICATION IS A FIRST-CLASS REQUIREMENT,\n   not a nice-to-have. ⚠️ Ensembles, conformal prediction, Gaussian\n   processes on fingerprints\n⚠️ THE METRIC THAT MATTERS  ⚠️ how many DMTA cycles to reach the target\n   profile — not test-set R²\n```\n**⚠️ Self-driving labs and closed-loop automation** connect this to robotic synthesis and\nassay — ⚠️ **and the real bottleneck becomes assay throughput and reliability, not\nprediction.**\n\n---\n\n## §17. ⚠️ Benchmarks\n\n**⚠️ Common ones**: **MoleculeNet, Therapeutics Data Commons (TDC), DUD-E and LIT-PCBA\n(virtual screening), CASF and PDBbind (scoring), GuacaMol and MOSES (generative), and\npolaris-style curated benchmarks.**\n> **⚠️ GOTCHA — most published benchmark improvements do not transfer, and there are\n> documented structural reasons.** ⚠️ **MoleculeNet's random splits are known to inflate\n> results (§7 → `drugdev-representation-cheminformatics-data-quality-and-leakage`); DUD-E's decoys are separable from actives by trivial properties, so models\n> learn the decoy generation procedure; PDBbind has train-test protein overlap; and\n> generative benchmarks reward distributional metrics that don't correspond to usefulness.**\n> **⚠️ Independent reanalyses have repeatedly found that simple baselines — random forest\n> on ECFP fingerprints — match or beat elaborate architectures once splits are fixed.**\n> **⚠️ Always run that baseline. If your model doesn't beat it, you have learned something\n> important.**\n\n---\n\n## §18. ⚠️ Validation That Means Something\n\n```\n⚠️ RETROSPECTIVE  time split (§7); ⚠️ external test set from a\n   different lab; performance stratified by applicability domain\n⚠️ PROSPECTIVE  ⚠️ THE ONLY REAL TEST — predict, then synthesize and\n   assay, and report the outcome including failures\n⚠️ THE COMPARISON THAT MATTERS  not \"better than random\" but\n   ⚠️ \"better than a medicinal chemist\" or \"better than the cheap baseline\"\n⚠️ CALIBRATION  are the confidence estimates honest? (§16)\n⚠️ ERROR ANALYSIS  where does it fail, and is that a systematic class?\n```\n**⚠️ The publication asymmetry to be aware of**: ⚠️ **prospective validations that failed\nare rarely published, so the visible literature overstates the field.** ⚠️ **This is the\nsame problem as §24.1 → `drugdev-reference` at a smaller scale.**\n\n---\n\n# PART V — ENGINEERING\n",{"name":5,"description":7},"develop","ae32736a1aabddabca5d2c96d665aa0ebc7d8700","src\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fmedicine-design-software-development\u002Fskills\u002Fdrugdev-qsar-admet-generative-models-and-validation",3,"https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey","https:\u002F\u002Fthe-vibey-project.github.io\u002Fvibey\u002F","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project","https:\u002F\u002Fvibewithadam.matthewsteinberger.com","freedom for software engineers","2026-11-15 05:47:42",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add the-vibey-project\u002Fvibey --skill drugdev-qsar-admet-generative-models-and-validation",{"tool":40,"command":41},"bunx","bunx skills add the-vibey-project\u002Fvibey --skill drugdev-qsar-admet-generative-models-and-validation",{"tool":43,"command":44},"pnpm","pnpm dlx skills add the-vibey-project\u002Fvibey --skill drugdev-qsar-admet-generative-models-and-validation","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fmedicine-design-software-development\u002Fskills\u002Fdrugdev-qsar-admet-generative-models-and-validation\u002FSKILL.md",true,[],[49,56,63,70,77,84],{"skillNo":50,"name":51,"nameZh":6,"description":52,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":53,"categories":54,"keywords":55,"publishedAt":19},"01M7AT1FQJAE1FQYAMDTF490Y4","white-paper-writing","Use when writing white papers, technical reports, policy documents, thought leadership pieces, or any long-form authoritative document. Triggers on requests to write, structure, outline, or review white papers, technical reports, position papers, policy briefs, or evidence-driven arguments. Also use for B2B content marketing strategy, technical documentation requiring authority, academic-style reports, or any document that must educate while persuading. Use when someone asks \"how should I structure this\" for a long-form document.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fwhite-paper-writing\u002FSKILL.md",[],[],{"skillNo":57,"name":58,"nameZh":6,"description":59,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":19},"01M7AT16BJRJJKK2MASXRZS76G","prose-technical-mechanics","Use when editing or writing prose at a technical\u002Fmechanical level: choosing between Anglo-Saxon and Latinate diction, placing information in topic vs stress positions, writing periodic vs cumulative sentences, handling sentence rhythm, punctuation as logic, transitions, paragraph structure, or phonesthetics. Use when someone wants to understand WHY prose works or doesn't, not just surface-level style advice. Triggers on word choice, sentence structure, emphasis, rhythm, diction, or any deep writing mechanics question.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fprose-technical-mechanics\u002FSKILL.md",[],[],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":19},"01M7AT0W9HT0ZZ9Y287NW99MR3","prose-mechanics","Use when writing, editing, or analyzing prose style; when trying to avoid generic AI-sounding writing; when applying specific prose styles such as journalistic, academic, narrative, KJV biblical, Hemingway plain, Latinate\u002FJamesian, Orwellian, Augustan, Victorian, Early Modern (Bacon, Browne, Milton), or free indirect discourse; or when coaching someone on writing voice and mechanics. Triggers on any request to improve, analyze, or write in a specific prose style, to fix \"AI-sounding\" writing, to make prose more vivid or concrete, to vary sentence rhythm, or to understand how a particular writer achieves their effects.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fprose-mechanics\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7AT0KR635JX66X2X6F3E2CP","narrative-structure","Use when structuring stories, pitches, presentations, white papers, or any persuasive narrative; analyzing the hero's journey in fiction or non-fiction; applying StoryBrand for business communication; designing narrative arcs for thought leadership; or understanding Vogler, Snyder, Propp, or Campbell frameworks. Triggers on any narrative structure, storytelling, pitch structure, StoryBrand, hero's journey, three-act structure, Freytag, Save the Cat, narrative arc, or \"how should I structure this story\u002Fpitch\u002Fpresentation\" request.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fnarrative-structure\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M7AT0AKGW09NKKZDYBT8VRDE","legal-brief-writing","Use when writing legal briefs, motions, or any court filing; when analyzing judicial cognition and persuasion; when applying elite brief-writing techniques from Roberts, Clement, Kagan, Garner; or when structuring legal arguments for maximum persuasion. Also use for autistic\u002Fneurodivergent advocates who need to understand how their natural communication style maps to elite legal advocacy. Triggers on any legal writing, brief drafting, motion preparation, or legal argument structuring request.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Flegal-brief-writing\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7AT00VWPS4Q8B5VEMXX9NQT","journal-publishability-criteria","Use whenever a research paper, article or preprint is being written, revised, reviewed or published, and specifically before every publication of docs\u002Fpaper.md. Covers what legitimate journals actually gate on, in order of gating — the editor's desk screen (scope, novelty, package problems), what reviewers judge (methodological soundness, claims the evidence can carry, statistical and qualitative reporting), EQUATOR reporting guidelines, the ICMJE authorship criteria, COPE's view of text recycling, the near-universal generative-AI policy (no AI author, disclosure, author responsibility), TOP 2025 transparency standards, how to verify a journal is legitimate, submitting without an institution, and a pre-submission checklist. Triggers on requests to evaluate, review, prepare or submit a paper, to judge publishability, to write a declarations or data-availability section, to disclose AI use, or to check a journal.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fjournal-publishability-criteria\u002FSKILL.md",[],[]]