{"id":"2084262f-60b6-4bb5-8c47-4b7060977ba1","shortId":"FqRDqt","kind":"skill","title":"write-protocol","tagline":"IRB/ethics committee research protocol generator. Produces 4 core sections (Background, Study Design, Sample Size, Statistical Plan) with full prose, plus 6 skeleton sections with TODO markers for institution-specific content. Integrates outputs from design-study, calc-sample-siz","description":"# Write-Protocol Skill\n\nYou are helping a medical researcher draft a research protocol for IRB/ethics committee\nsubmission. This skill generates the scientific core of the protocol while providing\nstructured skeletons for institution-specific sections.\n\n## Scope\n\nThis skill generates **4 core sections** with full prose:\n1. Background and Rationale\n2. Study Design and Eligibility Criteria\n3. Sample Size Justification\n4. Statistical Analysis Plan\n\nThe remaining **6 sections** are provided as structured skeletons with `[TODO]` markers,\nbecause they vary significantly across institutions, countries, and regulatory frameworks.\n\n**Important**: This protocol is a STARTING POINT. Every institution has its own IRB\nsubmission form and requirements. The generated protocol must be adapted to your\ninstitution's specific format before submission.\n\n## Reference Files\n\n- **Protocol template**: `${CLAUDE_SKILL_DIR}/references/protocol_template.md` -- complete 10-section structure with formatting guidance\n- **Ethics checklist**: `${CLAUDE_SKILL_DIR}/references/ethics_checklist.md` -- jurisdiction-specific ethical requirements\n\nRead both reference files before generating a protocol draft.\n\n## Cross-Skill Integration\n\n- **Input from design-study**: Study design recommendations, analysis unit, comparator design, validation strategy\n- **Input from calc-sample-size**: `protocol/sample_size_justification.md` (canonical IRB-ready prose) + `protocol/sample_size_calc.{R,py}` (reproducible code). Embed `sample_size_justification.md` VERBATIM into Methods §Sample Size — do not rephrase numbers (per `~/.claude/rules/numerical-safety.md`).\n- **Input from search-lit**: Background references with verified citations\n- **Input from define-variables**: `variable_operationalization.md` — literature-grounded definitions, cutoffs, DB-variable mappings for the Methods section. **Precondition**: if the study is observational and no operationalization artifact exists, call `/define-variables` before drafting Methods. Do not invent phenotype/cutoff definitions from the data dictionary inside this skill.\n- **Pipeline position**: search-lit -> design-study -> calc-sample-size -> define-variables -> **write-protocol** -> manage-project\n\nWhen prior skill outputs are available, incorporate them directly. When they are not,\nprompt the user or call the relevant skill.\n\n---\n\n## Communication Rules\n\n- Communicate with the user in their preferred language.\n- Use English for all protocol content, statistical terminology, and medical terminology.\n- Be explicit about what is generated versus what requires institutional input.\n\n---\n\n## Required Inputs\n\nCollect all required inputs before generating. Ask one question at a time if information is missing.\n\n1. **Research question / hypothesis** -- specific, testable\n2. **Study type** -- retrospective cohort, prospective cohort, cross-sectional, RCT, diagnostic accuracy, case-control, case series\n3. **Target population** -- who, where, when\n4. **Primary outcome** -- what you are measuring\n5. **Secondary outcomes** (if any)\n\n## Optional Inputs (enhance quality if available)\n\n6. **design-study output** -- if `/design-study` was already run, load its recommendations\n7. **calc-sample-size output** -- if `/calc-sample-size` was already run, load its results and IRB text\n8. **Key references** -- DOIs or search terms for background section\n9. **Institution name** -- for header and ethics section guidance\n10. **Regulatory context** -- Korea (PIPA), US (HIPAA/Common Rule), EU (GDPR), other\n\n---\n\n## Protocol Structure -- 10 Sections\n\n### Core Sections (Fully Generated)\n\n#### Section 1: Background and Rationale (400-600 words)\n\nGenerate full prose covering:\n- **Clinical context**: disease burden, current practice, knowledge gap\n- **Literature support**: call `/search-lit` if key references are not provided; every citation must have a verified DOI or PMID\n- **Rationale**: why this study is needed, what it adds to existing evidence\n- **Research question**: clear statement of the hypothesis or research question\n\nDo not use bullet points in the output. Write in full paragraphs with logical flow from\nclinical context through knowledge gap to research question.\n\n#### Section 2: Study Design and Eligibility Criteria (300-500 words)\n\nGenerate full prose plus structured criteria lists:\n- **Study design** with justification (why this design answers this question)\n- **Setting**: single-center vs multi-center, institution description\n- **Study period**: start and end dates or planned duration\n- **Inclusion criteria** (numbered list)\n- **Exclusion criteria** (numbered list)\n\nIf design-study output is available, incorporate its recommendations on:\n- Analysis unit (patient vs lesion vs exam)\n- Comparator design\n- Validation strategy\n- Potential leakage risks and mitigations\n\n#### Section 3: Sample Size Justification (150-300 words)\n\n- If `protocol/sample_size_justification.md` exists (calc-sample-size output): embed VERBATIM. Do not rephrase numbers.\n- If not available: prompt the user to run `/calc-sample-size` first; only fall back to a basic justification if the user explicitly declines.\n- Must include: test type, expected effect size (with literature source), alpha level, power, attrition adjustment\n- Final statement: \"We plan to enroll N participants.\"\n\n#### Section 4: Statistical Analysis Plan (300-500 words)\n\nGenerate full prose covering:\n- **Descriptive statistics**: continuous variables as mean (SD) or median (IQR); categorical variables as count (%)\n- **Primary analysis**: statistical test, assumptions, handling of violations\n- **Secondary analyses**: pre-specified\n- **Subgroup analyses**: pre-specified, with interaction tests\n- **Missing data**: handling strategy (complete case, multiple imputation, sensitivity analysis)\n- **Software**: name and version (e.g., R 4.4.0, Python 3.12, SAS 9.4)\n- **Significance level**: two-sided alpha = 0.05 unless otherwise justified\n\n### Skeleton Sections (TODO Markers)\n\n#### Section 5: Study Title and Registration\n\n```\n[TODO: Full study title]\n[TODO: Short title / acronym]\n[TODO: Clinical trial registry number if applicable (e.g., ClinicalTrials.gov, CRIS)]\n[TODO: Protocol version number and date]\n```\n\n#### Section 6: Data Collection and Management\n\n```\n[TODO: List variables to be collected -- use your institution's case report form (CRF) template]\n[TODO: Data collection method (chart review / prospective forms / electronic extraction)]\n[TODO: Data storage and security measures (encrypted database, access controls)]\n[TODO: Quality assurance procedures (double data entry, range checks)]\n[TODO: Data retention period per institutional policy]\n```\n\n#### Section 7: Ethical Considerations\n\n```\n[TODO: IRB/Ethics committee name and expected submission date]\n[TODO: Informed consent process -- or justification for waiver]\n[TODO: Patient privacy and data protection measures]\n```\n\nInclude regulatory guidance by jurisdiction:\n- **Korea**: Personal Information Protection Act (PIPA), Bioethics and Safety Act\n- **United States**: HIPAA Privacy Rule, Common Rule (45 CFR 46)\n- **European Union**: GDPR Article 9 (health data), Clinical Trials Regulation (EU 536/2014)\n\n```\n[TODO: Confirm applicable regulations with your IRB office]\n```\n\nRefer to `${CLAUDE_SKILL_DIR}/references/ethics_checklist.md` for the full checklist.\n\n#### Section 8: Timeline and Milestones\n\n```\n[TODO: Adapt to your project schedule]\n\n| Phase | Activity                              | Duration    | Target Date |\n|-------|---------------------------------------|-------------|-------------|\n| 1     | IRB approval                          | [X] weeks   | [TODO]      |\n| 2     | Data collection / Patient enrollment  | [X] months  | [TODO]      |\n| 3     | Data cleaning and analysis            | [X] months  | [TODO]      |\n| 4     | Manuscript preparation                | [X] months  | [TODO]      |\n| 5     | Submission                            | --          | [TODO]      |\n```\n\n#### Section 9: Budget\n\n```\n[TODO: Use your institution's budget template]\n[TODO: Common cost categories below -- delete or add as needed]\n- Personnel (research coordinator, statistician)\n- Equipment and supplies\n- Software licenses\n- Statistical consultation\n- Publication fees (open access APC)\n- Patient compensation (if applicable)\n```\n\n#### Section 10: References\n\nGenerate a numbered reference list from:\n- Citations used in Section 1 (Background and Rationale)\n- Effect size sources from Section 3 (Sample Size Justification)\n- Any additional references from calc-sample-size output\n\nAll references must have verified DOIs or PMIDs. Mark any unverified references\nas `[UNVERIFIED - NEEDS MANUAL CHECK]`.\n\n---\n\n## Output Format\n\nGenerate a single markdown file: `protocol_draft.md`\n\nRequirements:\n- All 10 sections with clear numbering (1. through 10.)\n- Core sections (1-4) in full prose, no bullet points in body text\n- Skeleton sections (5-9) with `[TODO]` markers clearly visible\n- Word count targets noted in comments at the start of each core section\n- Institution name in the header if provided\n\nAfter generating, inform the user:\n1. Which sections are complete and ready for review\n2. Which `[TODO]` items require their input\n3. Recommended next steps (e.g., \"Fill in Section 5 title and registration, then adapt Section 7 to your IRB form\")\n\n---\n\n## Quality Checks\n\nBefore delivering the protocol:\n\n1. **Citation integrity**: Every reference has a DOI or PMID, or is marked `[UNVERIFIED]`\n2. **Internal consistency**: Sample size in Section 3 matches the analysis plan in Section 4\n3. **Design alignment**: Study type in Section 2 matches the statistical approach in Section 4\n4. **TODO completeness**: All institution-specific items have `[TODO]` markers\n5. **Word counts**: Core sections fall within target ranges\n6. **No AI patterns**: Avoid phrases like \"it is worth noting\", \"comprehensive\", \"plays a crucial role\"\n\n## Anti-Hallucination\n\n- **Never fabricate references.** All citations must be verified via `/search-lit` with confirmed DOI or PMID. Mark unverified references as `[UNVERIFIED - NEEDS MANUAL CHECK]`.\n- **Never invent clinical definitions, diagnostic criteria, or guideline recommendations.** If uncertain, flag with `[VERIFY]` and ask the user.\n- **Never fabricate numerical results** — compliance percentages, scores, effect sizes, or sample sizes must come from actual data or analysis output.\n- If a reporting guideline item, journal policy, or clinical standard is uncertain, state the uncertainty rather than guessing.","tags":["write","protocol","medsci","skills","aperivue","agent-skills","biostatistics","claude-code","claude-skills","clinical-research","diagnostic-accuracy","irb-protocol"],"capabilities":["skill","source-aperivue","skill-write-protocol","topic-agent-skills","topic-biostatistics","topic-claude-code","topic-claude-skills","topic-clinical-research","topic-diagnostic-accuracy","topic-irb-protocol","topic-literature-review","topic-manuscript","topic-medical-ai","topic-medical-research","topic-meta-analysis"],"categories":["medsci-skills"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/Aperivue/medsci-skills/write-protocol","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add Aperivue/medsci-skills","source_repo":"https://github.com/Aperivue/medsci-skills","install_from":"skills.sh"}},"qualityScore":"0.499","qualityRationale":"deterministic score 0.50 from registry signals: · indexed on github topic:agent-skills · 98 github stars · SKILL.md body (10,545 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-05-18T18:56:32.203Z","embedding":null,"createdAt":"2026-05-13T12:57:47.145Z","updatedAt":"2026-05-18T18:56:32.203Z","lastSeenAt":"2026-05-18T18:56:32.203Z","tsv":"'-300':682 '-4':1154 '-500':603,749 '-600':516 '-9':1167 '/.claude/rules/numerical-safety.md':244 '/calc-sample-size':462,706 '/define-variables':286 '/design-study':448 '/references/ethics_checklist.md':182,989 '/references/protocol_template.md':169 '/search-lit':533,1332 '0.05':817 '1':91,394,511,1010,1094,1148,1153,1198,1240 '10':171,491,504,1082,1143,1150 '150':681 '2':95,400,596,1016,1207,1254,1276 '3':101,418,677,1024,1103,1214,1261,1269 '3.12':808 '300':602,748 '4':10,85,105,424,744,1032,1268,1283,1284 '4.4.0':806 '400':515 '45':961 '46':963 '5':431,826,1038,1166,1222,1295 '536/2014':975 '6':24,111,442,856,1304 '7':455,913,1229 '8':472,995 '9':482,968,1042 '9.4':810 'access':894,1075 'accuraci':412 'acronym':838 'across':125 'act':948,953 'activ':1006 'actual':1379 'adapt':153,1000,1227 'add':557,1058 'addit':1108 'adjust':734 'ai':1306 'align':1271 'alpha':730,816 'alreadi':450,464 'analys':778,783 'analysi':107,209,660,746,770,799,1028,1264,1382 'answer':619 'anti':1321 'anti-hallucin':1320 'apc':1076 'applic':845,978,1080 'approach':1280 'approv':1012 'articl':967 'artifact':283 'ask':384,1361 'assumpt':773 'assur':898 'attrit':733 'avail':328,441,655,700 'avoid':1308 'back':710 'background':13,92,250,480,512,1095 'basic':713 'bioethic':950 'bodi':1162 'budget':1043,1049 'bullet':574,1159 'burden':525 'calc':42,218,311,457,688,1112 'calc-sample-s':217,310,456,687,1111 'calc-sample-siz':41 'call':285,340,532 'canon':222 'case':414,416,795,871 'case-control':413 'categor':765 'categori':1054 'center':625,629 'cfr':962 'chart':880 'check':904,1132,1235,1345 'checklist':178,993 'citat':254,541,1090,1241,1327 'claud':166,179,986 'clean':1026 'clear':563,1146,1171 'clinic':522,587,840,971,1348,1392 'clinicaltrials.gov':847 'code':231 'cohort':404,406 'collect':378,858,866,878,1018 'come':1377 'comment':1178 'committe':5,61,918 'common':959,1052 'communic':344,346 'compar':211,667 'compens':1078 'complet':170,794,1202,1286 'complianc':1368 'comprehens':1315 'confirm':977,1334 'consent':926 'consider':915 'consist':1256 'consult':1071 'content':34,359 'context':493,523,588 'continu':757 'control':415,895 'coordin':1063 'core':11,68,86,506,1151,1184,1298 'cost':1053 'count':768,1174,1297 'countri':127 'cover':521,754 'crf':874 'cris':848 'criteria':100,601,610,642,646,1351 'cross':198,408 'cross-sect':407 'cross-skil':197 'crucial':1318 'current':526 'cutoff':265 'data':297,791,857,877,887,901,906,936,970,1017,1025,1380 'databas':893 'date':637,854,923,1009 'db':267 'db-variabl':266 'declin':719 'defin':258,315 'define-vari':257,314 'definit':264,294,1349 'delet':1056 'deliv':1237 'descript':631,755 'design':15,39,97,204,207,212,308,444,598,613,618,651,668,1270 'design-studi':38,203,307,443,650 'diagnost':411,1350 'dictionari':298 'dir':168,181,988 'direct':331 'diseas':524 'doi':475,546,1121,1247,1335 'doubl':900 'draft':55,196,288 'durat':640,1007 'e.g':804,846,1218 'effect':725,1098,1371 'electron':884 'elig':99,600 'emb':232,692 'encrypt':892 'end':636 'english':355 'enhanc':438 'enrol':740,1020 'entri':902 'equip':1065 'ethic':177,186,488,914 'eu':499,974 'european':964 'everi':138,540,1243 'evid':560 'exam':666 'exclus':645 'exist':284,559,686 'expect':724,921 'explicit':366,718 'extract':885 'fabric':1324,1365 'fall':709,1300 'fee':1073 'file':163,191,1139 'fill':1219 'final':735 'first':707 'flag':1357 'flow':585 'form':145,873,883,1233 'format':159,175,1134 'framework':130 'full':21,89,519,581,606,752,832,992,1156 'fulli':508 'gap':529,591 'gdpr':500,966 'generat':8,65,84,149,193,370,383,509,518,605,751,1084,1135,1194 'ground':263 'guess':1401 'guidanc':176,490,941 'guidelin':1353,1387 'hallucin':1322 'handl':774,792 'header':486,1190 'health':969 'help':51 'hipaa':956 'hipaa/common':497 'hypothesi':397,567 'import':131 'imput':797 'includ':721,939 'inclus':641 'incorpor':329,656 'inform':391,925,946,1195 'input':201,215,245,255,375,377,381,437,1213 'insid':299 'institut':32,78,126,139,156,374,483,630,869,910,1047,1186,1289 'institution-specif':31,77,1288 'integr':35,200,1242 'interact':788 'intern':1255 'invent':292,1347 'iqr':764 'irb':143,224,470,982,1011,1232 'irb-readi':223 'irb/ethics':4,60,917 'item':1210,1291,1388 'journal':1389 'jurisdict':184,943 'jurisdiction-specif':183 'justif':104,615,680,714,929,1106 'justifi':820 'key':473,535 'knowledg':528,590 'korea':494,944 'languag':353 'leakag':672 'lesion':664 'level':731,812 'licens':1069 'like':1310 'list':611,644,648,862,1088 'lit':249,306 'literatur':262,530,728 'literature-ground':261 'load':452,466 'logic':584 'manag':321,860 'manage-project':320 'manual':1131,1344 'manuscript':1033 'map':269 'mark':1124,1252,1338 'markdown':1138 'marker':29,120,824,1170,1294 'match':1262,1277 'mean':760 'measur':430,891,938 'median':763 'medic':53,363 'method':236,272,289,879 'mileston':998 'miss':393,790 'mitig':675 'month':1022,1030,1036 'multi':628 'multi-cent':627 'multipl':796 'must':151,542,720,1118,1328,1376 'n':741 'name':484,801,919,1187 'need':554,1060,1130,1343 'never':1323,1346,1364 'next':1216 'note':1176,1314 'number':242,643,647,697,843,852,1086,1147 'numer':1366 'observ':279 'offic':983 'one':385 'open':1074 'operation':282 'option':436 'otherwis':819 'outcom':426,433 'output':36,326,446,460,578,653,691,1115,1133,1383 'paragraph':582 'particip':742 'patient':662,933,1019,1077 'pattern':1307 'per':243,909 'percentag':1369 'period':633,908 'person':945 'personnel':1061 'phase':1005 'phenotype/cutoff':293 'phrase':1309 'pipa':495,949 'pipelin':302 'plan':19,108,639,738,747,1265 'play':1316 'plus':23,608 'pmid':548,1123,1249,1337 'point':137,575,1160 'polici':911,1390 'popul':420 'posit':303 'potenti':671 'power':732 'practic':527 'pre':780,785 'pre-specifi':779,784 'precondit':274 'prefer':352 'prepar':1034 'primari':425,769 'prior':324 'privaci':934,957 'procedur':899 'process':927 'produc':9 'project':322,1003 'prompt':336,701 'prose':22,90,226,520,607,753,1157 'prospect':405,882 'protect':937,947 'protocol':3,7,47,58,71,133,150,164,195,319,358,502,850,1239 'protocol/sample_size_calc':227 'protocol/sample_size_justification.md':221,685 'protocol_draft.md':1140 'provid':73,114,539,1192 'public':1072 'py':229 'python':807 'qualiti':439,897,1234 'question':386,396,562,570,594,621 'r':228,805 'rang':903,1303 'rather':1399 'rational':94,514,549,1097 'rct':410 'read':188 'readi':225,1204 'recommend':208,454,658,1215,1354 'refer':162,190,251,474,536,984,1083,1087,1109,1117,1127,1244,1325,1340 'registr':830,1225 'registri':842 'regul':973,979 'regulatori':129,492,940 'relev':342 'remain':110 'rephras':241,696 'report':872,1386 'reproduc':230 'requir':147,187,373,376,380,1141,1211 'research':6,54,57,395,561,569,593,1062 'result':468,1367 'retent':907 'retrospect':403 'review':881,1206 'risk':673 'role':1319 'rule':345,498,958,960 'run':451,465,705 'safeti':952 'sampl':16,43,102,219,237,312,458,678,689,1104,1113,1257,1374 'sample_size_justification.md':233 'sas':809 'schedul':1004 'scientif':67 'scope':81 'score':1370 'sd':761 'search':248,305,477 'search-lit':247,304 'secondari':432,777 'section':12,26,80,87,112,172,273,409,481,489,505,507,510,595,676,743,822,825,855,912,994,1041,1081,1093,1102,1144,1152,1165,1185,1200,1221,1228,1260,1267,1275,1282,1299 'secur':890 'sensit':798 'seri':417 'set':622 'short':836 'side':815 'signific':124,811 'singl':624,1137 'single-cent':623 'siz':44 'size':17,103,220,238,313,459,679,690,726,1099,1105,1114,1258,1372,1375 'skeleton':25,75,117,821,1164 'skill':48,64,83,167,180,199,301,325,343,987 'skill-write-protocol' 'softwar':800,1068 'sourc':729,1100 'source-aperivue' 'specif':33,79,158,185,398,1290 'specifi':781,786 'standard':1393 'start':136,634,1181 'state':955,1396 'statement':564,736 'statist':18,106,360,745,756,771,1070,1279 'statistician':1064 'step':1217 'storag':888 'strategi':214,670,793 'structur':74,116,173,503,609 'studi':14,40,96,205,206,277,309,401,445,552,597,612,632,652,827,833,1272 'subgroup':782 'submiss':62,144,161,922,1039 'suppli':1067 'support':531 'target':419,1008,1175,1302 'templat':165,875,1050 'term':478 'terminolog':361,364 'test':722,772,789 'testabl':399 'text':471,1163 'time':389 'timelin':996 'titl':828,834,837,1223 'todo':28,119,823,831,835,839,849,861,876,886,896,905,916,924,932,976,999,1015,1023,1031,1037,1040,1044,1051,1169,1209,1285,1293 'topic-agent-skills' 'topic-biostatistics' 'topic-claude-code' 'topic-claude-skills' 'topic-clinical-research' 'topic-diagnostic-accuracy' 'topic-irb-protocol' 'topic-literature-review' 'topic-manuscript' 'topic-medical-ai' 'topic-medical-research' 'topic-meta-analysis' 'trial':841,972 'two':814 'two-sid':813 'type':402,723,1273 'uncertain':1356,1395 'uncertainti':1398 'union':965 'unit':210,661,954 'unless':818 'unverifi':1126,1129,1253,1339,1342 'us':496 'use':354,573,867,1045,1091 'user':338,349,703,717,1197,1363 'valid':213,669 'vari':123 'variabl':259,268,316,758,766,863 'variable_operationalization.md':260 'verbatim':234,693 'verifi':253,545,1120,1330,1359 'version':803,851 'versus':371 'via':1331 'violat':776 'visibl':1172 'vs':626,663,665 'waiver':931 'week':1014 'within':1301 'word':517,604,683,750,1173,1296 'worth':1313 'write':2,46,318,579 'write-protocol':1,45,317 'x':1013,1021,1029,1035","prices":[{"id":"1a0e22e1-bbdc-47bd-9c40-ba912c7492c0","listingId":"2084262f-60b6-4bb5-8c47-4b7060977ba1","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"Aperivue","category":"medsci-skills","install_from":"skills.sh"},"createdAt":"2026-05-13T12:57:47.145Z"}],"sources":[{"listingId":"2084262f-60b6-4bb5-8c47-4b7060977ba1","source":"github","sourceId":"Aperivue/medsci-skills/write-protocol","sourceUrl":"https://github.com/Aperivue/medsci-skills/tree/main/skills/write-protocol","isPrimary":false,"firstSeenAt":"2026-05-13T12:57:47.145Z","lastSeenAt":"2026-05-18T18:56:32.203Z"}],"details":{"listingId":"2084262f-60b6-4bb5-8c47-4b7060977ba1","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"Aperivue","slug":"write-protocol","github":{"repo":"Aperivue/medsci-skills","stars":98,"topics":["agent-skills","biostatistics","claude-code","claude-skills","clinical-research","diagnostic-accuracy","irb-protocol","literature-review","manuscript","medical-ai","medical-research","meta-analysis","physician-researcher","prisma","pubmed","radiology","reporting-guidelines","strobe","systematic-review","tripod-ai"],"license":"other","html_url":"https://github.com/Aperivue/medsci-skills","pushed_at":"2026-05-17T20:50:52Z","description":"Claude Code skills for medical research — literature search, reporting guidelines, statistical analysis, publication figures. Built by a physician-researcher, tested on real publications. MIT licensed.","skill_md_sha":"2e679f883db8f759d8c0f7c4731e469ec3a82668","skill_md_path":"skills/write-protocol/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/Aperivue/medsci-skills/tree/main/skills/write-protocol"},"layout":"multi","source":"github","category":"medsci-skills","frontmatter":{"name":"write-protocol","description":"IRB/ethics committee research protocol generator. Produces 4 core sections (Background, Study Design, Sample Size, Statistical Plan) with full prose, plus 6 skeleton sections with TODO markers for institution-specific content. Integrates outputs from design-study, calc-sample-size, and search-lit."},"skills_sh_url":"https://skills.sh/Aperivue/medsci-skills/write-protocol"},"updatedAt":"2026-05-18T18:56:32.203Z"}}