[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-nvidia-cohort-compare":3,"mdc--ll6mhd-key":34,"related-org-nvidia-cohort-compare":743,"related-repo-nvidia-cohort-compare":903},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":23,"repoUrl":24,"updatedAt":25,"license":26,"forks":27,"topics":28,"repo":29,"sourceUrl":32,"mdContent":33},"cohort-compare","analyze patient cohorts from FHIR endpoints","Analyze a cohort of patients from FHIR endpoints to find care gaps and patterns. Use when asked to compare patients, find quality gaps, or analyze a population.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},"nvidia","NVIDIA","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fnvidia.png",[12,16,19,20],{"name":13,"slug":14,"type":15},"Healthcare","healthcare","tag",{"name":17,"slug":18,"type":15},"Data Analysis","data-analysis",{"name":9,"slug":8,"type":15},{"name":21,"slug":22,"type":15},"FHIR","fhir",1144,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fdgx-spark-playbooks","2026-07-14T05:35:51.543095",null,249,[],{"repoUrl":24,"stars":23,"forks":27,"topics":30,"description":31},[],"Collection of step-by-step playbooks for setting up AI\u002FML workloads on NVIDIA DGX Spark devices with Blackwell architecture.","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fdgx-spark-playbooks\u002Ftree\u002FHEAD\u002Fnvidia\u002Fstation-healthcare-agent\u002Fassets\u002Fskills\u002Fcohort-compare","---\nname: cohort-compare\ndescription: Analyze a cohort of patients from FHIR endpoints to find care gaps and patterns. Use when asked to compare patients, find quality gaps, or analyze a population.\nmetadata:\n  openclaw:\n    requires:\n      bins: [\"python3\"]\n---\n\nAnalyze a patient cohort: $ARGUMENTS\n\nUse your fhir-basics skill to query FHIR endpoints. Use your clinical-knowledge skill to identify care gaps and apply correct thresholds. Use your analysis-methods skill to write correct Python analysis code.\n\n## Execution Rules\n\n- Do NOT explore the workspace or list files. Begin the analysis immediately.\n- Write ONE Python script that does everything: FHIR queries, analysis, chart, and summary.\n- Write the script correctly the first time. Do NOT write a draft and then edit it.\n- Run it with `python` (NOT `python3`).\n- All HTTP calls must use `subprocess.run([\"curl\", \"-sf\", \"--max-time\", \"30\", url], capture_output=True, text=True)` -- the `requests` library does NOT work through the sandbox proxy. See the fhir-basics skill for the `fhir_get` helper pattern.\n- Save the script to `\u002Ftmp\u002F\u003Cname>.py`, run it once, interpret the output.\n\n## Steps\n\n1. **Identify the cohort** -- Query `GET \u002FCondition?code={snomed_code}&_count=200` and follow pagination links to get all matching Condition resources. Extract unique patient IDs from `entry[].resource.subject.reference`. Report the cohort size.\n\n2. **Pull clinical data in BATCHED queries (do NOT loop per-patient)**:\n   - **Lab values**: Use `get_latest_labs_batch(loinc_code, patient_ids)` to fetch ALL observations for the LOINC code in one call and filter client-side. This queries `GET \u002FObservation?code={loinc_code}&_count=500&_sort=-date` without a patient filter, then builds a dict keyed by patient ID. Handle both `valueQuantity` (numeric) and `valueString` (text) formats. For blood pressure, query the BP panel code `85354-9` in batch and parse components.\n   - **Medications**: Use `get_all_medications_batch(patient_ids)` to fetch `GET \u002FMedicationRequest?status=active&_count=500` in one call, then filter to cohort patients client-side.\n   - **NEVER write a `for pid in patient_ids:` loop that makes FHIR HTTP calls inside the loop.** The sandbox proxy adds 1-3s latency per call. With 24 patients x 4 LOINC codes = 96 calls = 5+ minutes. Batching brings this to 4-6 total calls = 30 seconds.\n\n3. **Build a pandas DataFrame** with one row per patient:\n   - `patient_id` (string)\n   - `{lab_name}` (float or None)\n   - `lab_date` (string)\n   - `on_target_med` (boolean -- True if the patient is on the specified medication class)\n   - `medications` (comma-separated string of all active med names)\n   - `med_count` (int)\n\n4. **Data quality check**:\n   - Report how many patients have the lab recorded vs. missing\n   - If > 30% missing, flag as a data quality issue but continue analysis\n   - If \u003C 5 patients have data, warn that the sample is too small for meaningful statistics\n\n5. **Identify care gap patients**: Apply the threshold and medication check:\n   - Gap = lab value exceeds threshold AND patient is NOT on the specified medication class\n   - Report: total with condition, total with lab data, total above threshold, total in care gap\n   - Compute gap rate as percentage of patients with lab data (not total cohort)\n\n6. **Generate visualization**:\n   - Histogram of the lab value distribution with threshold line\n   - Use NVIDIA dark theme: `plt.style.use('dark_background')`, primary color `#76B900`, background `#1a1a1a`\n   - Annotate with sample size (N = ...) and gap count\n   - Save as PNG with `dpi=150`\n\n7. **Write a plain-English summary** including:\n   - Cohort size and data completeness\n   - Distribution statistics (mean, median, range)\n   - Gap patient count and percentage with absolute numbers\n   - Comparison to relevant CMS quality measure if applicable\n   - Any notable patterns (e.g., \"All gap patients had no medications recorded at all\")\n\n8. **Disclaimer**: \"This analysis is for research and operational purposes. Clinical decisions should be made by qualified clinicians.\"\n\n## Example: Diabetes Gap Analysis (CMS122)\n\nCondition: Type 2 Diabetes (SNOMED 44054006)\nLab: HbA1c (LOINC 4548-4)\nThreshold: > 9.0%\nGap medication: insulin or GLP-1 agonist\nQuality measure: CMS122v12 (poor glycemic control)\n\n```python\nINSULIN_AND_GLP1 = [\"insulin\", \"liraglutide\", \"semaglutide\", \"dulaglutide\",\n                     \"exenatide\", \"tirzepatide\", \"victoza\", \"ozempic\",\n                     \"trulicity\", \"byetta\", \"mounjaro\", \"rybelsus\"]\n\ndef is_on_insulin_or_glp1(med_list):\n    med_lower = [m.lower() for m in med_list]\n    return any(drug in med_text for drug in INSULIN_AND_GLP1 for med_text in med_lower)\n```\n\n## Example: Hypertension Gap Analysis (CMS165)\n\nCondition: Essential Hypertension (SNOMED 38341003)\nLab: Systolic BP (LOINC 8480-6) -- **use component Observation pattern**\nThreshold: >= 140 mmHg\nGap medication: any antihypertensive\nQuality measure: CMS165v12 (controlling high blood pressure)\n\nNote: Use `get_latest_bp()` from the analysis-methods skill to handle both BP panel (85354-9) and standalone systolic Observations.\n\n```python\nANTIHYPERTENSIVES = [\"lisinopril\", \"enalapril\", \"ramipril\", \"benazepril\",\n    \"losartan\", \"valsartan\", \"irbesartan\", \"olmesartan\", \"telmisartan\",\n    \"amlodipine\", \"nifedipine\", \"diltiazem\",\n    \"metoprolol\", \"atenolol\", \"carvedilol\", \"bisoprolol\",\n    \"hydrochlorothiazide\", \"hctz\", \"chlorthalidone\",\n    \"furosemide\", \"spironolactone\"]\n\ndef is_on_antihypertensive(med_list):\n    med_lower = [m.lower() for m in med_list]\n    return any(drug in med_text for drug in ANTIHYPERTENSIVES for med_text in med_lower)\n```\n",{"data":35,"body":41},{"name":4,"description":6,"metadata":36},{"openclaw":37},{"requires":38},{"bins":39},[40],"python3",{"type":42,"children":43},"root",[44,52,57,64,147,153,531,537,542,618,624,636,649,737],{"type":45,"tag":46,"props":47,"children":48},"element","p",{},[49],{"type":50,"value":51},"text","Analyze a patient cohort: $ARGUMENTS",{"type":45,"tag":46,"props":53,"children":54},{},[55],{"type":50,"value":56},"Use your fhir-basics skill to query FHIR endpoints. Use your clinical-knowledge skill to identify care gaps and apply correct thresholds. Use your analysis-methods skill to write correct Python analysis code.",{"type":45,"tag":58,"props":59,"children":61},"h2",{"id":60},"execution-rules",[62],{"type":50,"value":63},"Execution Rules",{"type":45,"tag":65,"props":66,"children":67},"ul",{},[68,74,79,84,105,134],{"type":45,"tag":69,"props":70,"children":71},"li",{},[72],{"type":50,"value":73},"Do NOT explore the workspace or list files. Begin the analysis immediately.",{"type":45,"tag":69,"props":75,"children":76},{},[77],{"type":50,"value":78},"Write ONE Python script that does everything: FHIR queries, analysis, chart, and summary.",{"type":45,"tag":69,"props":80,"children":81},{},[82],{"type":50,"value":83},"Write the script correctly the first time. Do NOT write a draft and then edit it.",{"type":45,"tag":69,"props":85,"children":86},{},[87,89,96,98,103],{"type":50,"value":88},"Run it with ",{"type":45,"tag":90,"props":91,"children":93},"code",{"className":92},[],[94],{"type":50,"value":95},"python",{"type":50,"value":97}," (NOT ",{"type":45,"tag":90,"props":99,"children":101},{"className":100},[],[102],{"type":50,"value":40},{"type":50,"value":104},").",{"type":45,"tag":69,"props":106,"children":107},{},[108,110,116,118,124,126,132],{"type":50,"value":109},"All HTTP calls must use ",{"type":45,"tag":90,"props":111,"children":113},{"className":112},[],[114],{"type":50,"value":115},"subprocess.run([\"curl\", \"-sf\", \"--max-time\", \"30\", url], capture_output=True, text=True)",{"type":50,"value":117}," -- the ",{"type":45,"tag":90,"props":119,"children":121},{"className":120},[],[122],{"type":50,"value":123},"requests",{"type":50,"value":125}," library does NOT work through the sandbox proxy. See the fhir-basics skill for the ",{"type":45,"tag":90,"props":127,"children":129},{"className":128},[],[130],{"type":50,"value":131},"fhir_get",{"type":50,"value":133}," helper pattern.",{"type":45,"tag":69,"props":135,"children":136},{},[137,139,145],{"type":50,"value":138},"Save the script to ",{"type":45,"tag":90,"props":140,"children":142},{"className":141},[],[143],{"type":50,"value":144},"\u002Ftmp\u002F\u003Cname>.py",{"type":50,"value":146},", run it once, interpret the output.",{"type":45,"tag":58,"props":148,"children":150},{"id":149},"steps",[151],{"type":50,"value":152},"Steps",{"type":45,"tag":154,"props":155,"children":156},"ol",{},[157,184,290,368,395,423,483,521],{"type":45,"tag":69,"props":158,"children":159},{},[160,166,168,174,176,182],{"type":45,"tag":161,"props":162,"children":163},"strong",{},[164],{"type":50,"value":165},"Identify the cohort",{"type":50,"value":167}," -- Query ",{"type":45,"tag":90,"props":169,"children":171},{"className":170},[],[172],{"type":50,"value":173},"GET \u002FCondition?code={snomed_code}&_count=200",{"type":50,"value":175}," and follow pagination links to get all matching Condition resources. Extract unique patient IDs from ",{"type":45,"tag":90,"props":177,"children":179},{"className":178},[],[180],{"type":50,"value":181},"entry[].resource.subject.reference",{"type":50,"value":183},". Report the cohort size.",{"type":45,"tag":69,"props":185,"children":186},{},[187,192,194],{"type":45,"tag":161,"props":188,"children":189},{},[190],{"type":50,"value":191},"Pull clinical data in BATCHED queries (do NOT loop per-patient)",{"type":50,"value":193},":",{"type":45,"tag":65,"props":195,"children":196},{},[197,247,272],{"type":45,"tag":69,"props":198,"children":199},{},[200,205,207,213,215,221,223,229,231,237,239,245],{"type":45,"tag":161,"props":201,"children":202},{},[203],{"type":50,"value":204},"Lab values",{"type":50,"value":206},": Use ",{"type":45,"tag":90,"props":208,"children":210},{"className":209},[],[211],{"type":50,"value":212},"get_latest_labs_batch(loinc_code, patient_ids)",{"type":50,"value":214}," to fetch ALL observations for the LOINC code in one call and filter client-side. This queries ",{"type":45,"tag":90,"props":216,"children":218},{"className":217},[],[219],{"type":50,"value":220},"GET \u002FObservation?code={loinc_code}&_count=500&_sort=-date",{"type":50,"value":222}," without a patient filter, then builds a dict keyed by patient ID. Handle both ",{"type":45,"tag":90,"props":224,"children":226},{"className":225},[],[227],{"type":50,"value":228},"valueQuantity",{"type":50,"value":230}," (numeric) and ",{"type":45,"tag":90,"props":232,"children":234},{"className":233},[],[235],{"type":50,"value":236},"valueString",{"type":50,"value":238}," (text) formats. For blood pressure, query the BP panel code ",{"type":45,"tag":90,"props":240,"children":242},{"className":241},[],[243],{"type":50,"value":244},"85354-9",{"type":50,"value":246}," in batch and parse components.",{"type":45,"tag":69,"props":248,"children":249},{},[250,255,256,262,264,270],{"type":45,"tag":161,"props":251,"children":252},{},[253],{"type":50,"value":254},"Medications",{"type":50,"value":206},{"type":45,"tag":90,"props":257,"children":259},{"className":258},[],[260],{"type":50,"value":261},"get_all_medications_batch(patient_ids)",{"type":50,"value":263}," to fetch ",{"type":45,"tag":90,"props":265,"children":267},{"className":266},[],[268],{"type":50,"value":269},"GET \u002FMedicationRequest?status=active&_count=500",{"type":50,"value":271}," in one call, then filter to cohort patients client-side.",{"type":45,"tag":69,"props":273,"children":274},{},[275,288],{"type":45,"tag":161,"props":276,"children":277},{},[278,280,286],{"type":50,"value":279},"NEVER write a ",{"type":45,"tag":90,"props":281,"children":283},{"className":282},[],[284],{"type":50,"value":285},"for pid in patient_ids:",{"type":50,"value":287}," loop that makes FHIR HTTP calls inside the loop.",{"type":50,"value":289}," The sandbox proxy adds 1-3s latency per call. With 24 patients x 4 LOINC codes = 96 calls = 5+ minutes. Batching brings this to 4-6 total calls = 30 seconds.",{"type":45,"tag":69,"props":291,"children":292},{},[293,298,300],{"type":45,"tag":161,"props":294,"children":295},{},[296],{"type":50,"value":297},"Build a pandas DataFrame",{"type":50,"value":299}," with one row per patient:",{"type":45,"tag":65,"props":301,"children":302},{},[303,314,325,335,346,357],{"type":45,"tag":69,"props":304,"children":305},{},[306,312],{"type":45,"tag":90,"props":307,"children":309},{"className":308},[],[310],{"type":50,"value":311},"patient_id",{"type":50,"value":313}," (string)",{"type":45,"tag":69,"props":315,"children":316},{},[317,323],{"type":45,"tag":90,"props":318,"children":320},{"className":319},[],[321],{"type":50,"value":322},"{lab_name}",{"type":50,"value":324}," (float or None)",{"type":45,"tag":69,"props":326,"children":327},{},[328,334],{"type":45,"tag":90,"props":329,"children":331},{"className":330},[],[332],{"type":50,"value":333},"lab_date",{"type":50,"value":313},{"type":45,"tag":69,"props":336,"children":337},{},[338,344],{"type":45,"tag":90,"props":339,"children":341},{"className":340},[],[342],{"type":50,"value":343},"on_target_med",{"type":50,"value":345}," (boolean -- True if the patient is on the specified medication class)",{"type":45,"tag":69,"props":347,"children":348},{},[349,355],{"type":45,"tag":90,"props":350,"children":352},{"className":351},[],[353],{"type":50,"value":354},"medications",{"type":50,"value":356}," (comma-separated string of all active med names)",{"type":45,"tag":69,"props":358,"children":359},{},[360,366],{"type":45,"tag":90,"props":361,"children":363},{"className":362},[],[364],{"type":50,"value":365},"med_count",{"type":50,"value":367}," (int)",{"type":45,"tag":69,"props":369,"children":370},{},[371,376,377],{"type":45,"tag":161,"props":372,"children":373},{},[374],{"type":50,"value":375},"Data quality check",{"type":50,"value":193},{"type":45,"tag":65,"props":378,"children":379},{},[380,385,390],{"type":45,"tag":69,"props":381,"children":382},{},[383],{"type":50,"value":384},"Report how many patients have the lab recorded vs. missing",{"type":45,"tag":69,"props":386,"children":387},{},[388],{"type":50,"value":389},"If > 30% missing, flag as a data quality issue but continue analysis",{"type":45,"tag":69,"props":391,"children":392},{},[393],{"type":50,"value":394},"If \u003C 5 patients have data, warn that the sample is too small for meaningful statistics",{"type":45,"tag":69,"props":396,"children":397},{},[398,403,405],{"type":45,"tag":161,"props":399,"children":400},{},[401],{"type":50,"value":402},"Identify care gap patients",{"type":50,"value":404},": Apply the threshold and medication check:",{"type":45,"tag":65,"props":406,"children":407},{},[408,413,418],{"type":45,"tag":69,"props":409,"children":410},{},[411],{"type":50,"value":412},"Gap = lab value exceeds threshold AND patient is NOT on the specified medication class",{"type":45,"tag":69,"props":414,"children":415},{},[416],{"type":50,"value":417},"Report: total with condition, total with lab data, total above threshold, total in care gap",{"type":45,"tag":69,"props":419,"children":420},{},[421],{"type":50,"value":422},"Compute gap rate as percentage of patients with lab data (not total cohort)",{"type":45,"tag":69,"props":424,"children":425},{},[426,431,432],{"type":45,"tag":161,"props":427,"children":428},{},[429],{"type":50,"value":430},"Generate visualization",{"type":50,"value":193},{"type":45,"tag":65,"props":433,"children":434},{},[435,440,467,472],{"type":45,"tag":69,"props":436,"children":437},{},[438],{"type":50,"value":439},"Histogram of the lab value distribution with threshold line",{"type":45,"tag":69,"props":441,"children":442},{},[443,445,451,453,459,461],{"type":50,"value":444},"Use NVIDIA dark theme: ",{"type":45,"tag":90,"props":446,"children":448},{"className":447},[],[449],{"type":50,"value":450},"plt.style.use('dark_background')",{"type":50,"value":452},", primary color ",{"type":45,"tag":90,"props":454,"children":456},{"className":455},[],[457],{"type":50,"value":458},"#76B900",{"type":50,"value":460},", background ",{"type":45,"tag":90,"props":462,"children":464},{"className":463},[],[465],{"type":50,"value":466},"#1a1a1a",{"type":45,"tag":69,"props":468,"children":469},{},[470],{"type":50,"value":471},"Annotate with sample size (N = ...) and gap count",{"type":45,"tag":69,"props":473,"children":474},{},[475,477],{"type":50,"value":476},"Save as PNG with ",{"type":45,"tag":90,"props":478,"children":480},{"className":479},[],[481],{"type":50,"value":482},"dpi=150",{"type":45,"tag":69,"props":484,"children":485},{},[486,491,493],{"type":45,"tag":161,"props":487,"children":488},{},[489],{"type":50,"value":490},"Write a plain-English summary",{"type":50,"value":492}," including:",{"type":45,"tag":65,"props":494,"children":495},{},[496,501,506,511,516],{"type":45,"tag":69,"props":497,"children":498},{},[499],{"type":50,"value":500},"Cohort size and data completeness",{"type":45,"tag":69,"props":502,"children":503},{},[504],{"type":50,"value":505},"Distribution statistics (mean, median, range)",{"type":45,"tag":69,"props":507,"children":508},{},[509],{"type":50,"value":510},"Gap patient count and percentage with absolute numbers",{"type":45,"tag":69,"props":512,"children":513},{},[514],{"type":50,"value":515},"Comparison to relevant CMS quality measure if applicable",{"type":45,"tag":69,"props":517,"children":518},{},[519],{"type":50,"value":520},"Any notable patterns (e.g., \"All gap patients had no medications recorded at all\")",{"type":45,"tag":69,"props":522,"children":523},{},[524,529],{"type":45,"tag":161,"props":525,"children":526},{},[527],{"type":50,"value":528},"Disclaimer",{"type":50,"value":530},": \"This analysis is for research and operational purposes. Clinical decisions should be made by qualified clinicians.\"",{"type":45,"tag":58,"props":532,"children":534},{"id":533},"example-diabetes-gap-analysis-cms122",[535],{"type":50,"value":536},"Example: Diabetes Gap Analysis (CMS122)",{"type":45,"tag":46,"props":538,"children":539},{},[540],{"type":50,"value":541},"Condition: Type 2 Diabetes (SNOMED 44054006)\nLab: HbA1c (LOINC 4548-4)\nThreshold: > 9.0%\nGap medication: insulin or GLP-1 agonist\nQuality measure: CMS122v12 (poor glycemic control)",{"type":45,"tag":543,"props":544,"children":548},"pre",{"className":545,"code":546,"language":95,"meta":547,"style":547},"language-python shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","INSULIN_AND_GLP1 = [\"insulin\", \"liraglutide\", \"semaglutide\", \"dulaglutide\",\n                     \"exenatide\", \"tirzepatide\", \"victoza\", \"ozempic\",\n                     \"trulicity\", \"byetta\", \"mounjaro\", \"rybelsus\"]\n\ndef is_on_insulin_or_glp1(med_list):\n    med_lower = [m.lower() for m in med_list]\n    return any(drug in med_text for drug in INSULIN_AND_GLP1 for med_text in med_lower)\n","",[549],{"type":45,"tag":90,"props":550,"children":551},{"__ignoreMap":547},[552,563,572,581,591,600,609],{"type":45,"tag":553,"props":554,"children":557},"span",{"class":555,"line":556},"line",1,[558],{"type":45,"tag":553,"props":559,"children":560},{},[561],{"type":50,"value":562},"INSULIN_AND_GLP1 = [\"insulin\", \"liraglutide\", \"semaglutide\", \"dulaglutide\",\n",{"type":45,"tag":553,"props":564,"children":566},{"class":555,"line":565},2,[567],{"type":45,"tag":553,"props":568,"children":569},{},[570],{"type":50,"value":571},"                     \"exenatide\", \"tirzepatide\", \"victoza\", \"ozempic\",\n",{"type":45,"tag":553,"props":573,"children":575},{"class":555,"line":574},3,[576],{"type":45,"tag":553,"props":577,"children":578},{},[579],{"type":50,"value":580},"                     \"trulicity\", \"byetta\", \"mounjaro\", \"rybelsus\"]\n",{"type":45,"tag":553,"props":582,"children":584},{"class":555,"line":583},4,[585],{"type":45,"tag":553,"props":586,"children":588},{"emptyLinePlaceholder":587},true,[589],{"type":50,"value":590},"\n",{"type":45,"tag":553,"props":592,"children":594},{"class":555,"line":593},5,[595],{"type":45,"tag":553,"props":596,"children":597},{},[598],{"type":50,"value":599},"def is_on_insulin_or_glp1(med_list):\n",{"type":45,"tag":553,"props":601,"children":603},{"class":555,"line":602},6,[604],{"type":45,"tag":553,"props":605,"children":606},{},[607],{"type":50,"value":608},"    med_lower = [m.lower() for m in med_list]\n",{"type":45,"tag":553,"props":610,"children":612},{"class":555,"line":611},7,[613],{"type":45,"tag":553,"props":614,"children":615},{},[616],{"type":50,"value":617},"    return any(drug in med_text for drug in INSULIN_AND_GLP1 for med_text in med_lower)\n",{"type":45,"tag":58,"props":619,"children":621},{"id":620},"example-hypertension-gap-analysis-cms165",[622],{"type":50,"value":623},"Example: Hypertension Gap Analysis (CMS165)",{"type":45,"tag":46,"props":625,"children":626},{},[627,629,634],{"type":50,"value":628},"Condition: Essential Hypertension (SNOMED 38341003)\nLab: Systolic BP (LOINC 8480-6) -- ",{"type":45,"tag":161,"props":630,"children":631},{},[632],{"type":50,"value":633},"use component Observation pattern",{"type":50,"value":635},"\nThreshold: >= 140 mmHg\nGap medication: any antihypertensive\nQuality measure: CMS165v12 (controlling high blood pressure)",{"type":45,"tag":46,"props":637,"children":638},{},[639,641,647],{"type":50,"value":640},"Note: Use ",{"type":45,"tag":90,"props":642,"children":644},{"className":643},[],[645],{"type":50,"value":646},"get_latest_bp()",{"type":50,"value":648}," from the analysis-methods skill to handle both BP panel (85354-9) and standalone systolic Observations.",{"type":45,"tag":543,"props":650,"children":652},{"className":545,"code":651,"language":95,"meta":547,"style":547},"ANTIHYPERTENSIVES = [\"lisinopril\", \"enalapril\", \"ramipril\", \"benazepril\",\n    \"losartan\", \"valsartan\", \"irbesartan\", \"olmesartan\", \"telmisartan\",\n    \"amlodipine\", \"nifedipine\", \"diltiazem\",\n    \"metoprolol\", \"atenolol\", \"carvedilol\", \"bisoprolol\",\n    \"hydrochlorothiazide\", \"hctz\", \"chlorthalidone\",\n    \"furosemide\", \"spironolactone\"]\n\ndef is_on_antihypertensive(med_list):\n    med_lower = [m.lower() for m in med_list]\n    return any(drug in med_text for drug in ANTIHYPERTENSIVES for med_text in med_lower)\n",[653],{"type":45,"tag":90,"props":654,"children":655},{"__ignoreMap":547},[656,664,672,680,688,696,704,711,720,728],{"type":45,"tag":553,"props":657,"children":658},{"class":555,"line":556},[659],{"type":45,"tag":553,"props":660,"children":661},{},[662],{"type":50,"value":663},"ANTIHYPERTENSIVES = [\"lisinopril\", \"enalapril\", \"ramipril\", \"benazepril\",\n",{"type":45,"tag":553,"props":665,"children":666},{"class":555,"line":565},[667],{"type":45,"tag":553,"props":668,"children":669},{},[670],{"type":50,"value":671},"    \"losartan\", \"valsartan\", \"irbesartan\", \"olmesartan\", \"telmisartan\",\n",{"type":45,"tag":553,"props":673,"children":674},{"class":555,"line":574},[675],{"type":45,"tag":553,"props":676,"children":677},{},[678],{"type":50,"value":679},"    \"amlodipine\", \"nifedipine\", \"diltiazem\",\n",{"type":45,"tag":553,"props":681,"children":682},{"class":555,"line":583},[683],{"type":45,"tag":553,"props":684,"children":685},{},[686],{"type":50,"value":687},"    \"metoprolol\", \"atenolol\", \"carvedilol\", \"bisoprolol\",\n",{"type":45,"tag":553,"props":689,"children":690},{"class":555,"line":593},[691],{"type":45,"tag":553,"props":692,"children":693},{},[694],{"type":50,"value":695},"    \"hydrochlorothiazide\", \"hctz\", \"chlorthalidone\",\n",{"type":45,"tag":553,"props":697,"children":698},{"class":555,"line":602},[699],{"type":45,"tag":553,"props":700,"children":701},{},[702],{"type":50,"value":703},"    \"furosemide\", \"spironolactone\"]\n",{"type":45,"tag":553,"props":705,"children":706},{"class":555,"line":611},[707],{"type":45,"tag":553,"props":708,"children":709},{"emptyLinePlaceholder":587},[710],{"type":50,"value":590},{"type":45,"tag":553,"props":712,"children":714},{"class":555,"line":713},8,[715],{"type":45,"tag":553,"props":716,"children":717},{},[718],{"type":50,"value":719},"def is_on_antihypertensive(med_list):\n",{"type":45,"tag":553,"props":721,"children":723},{"class":555,"line":722},9,[724],{"type":45,"tag":553,"props":725,"children":726},{},[727],{"type":50,"value":608},{"type":45,"tag":553,"props":729,"children":731},{"class":555,"line":730},10,[732],{"type":45,"tag":553,"props":733,"children":734},{},[735],{"type":50,"value":736},"    return any(drug in med_text for drug in ANTIHYPERTENSIVES for med_text in med_lower)\n",{"type":45,"tag":738,"props":739,"children":740},"style",{},[741],{"type":50,"value":742},"html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"items":744,"total":902},[745,763,780,791,803,817,830,844,857,868,882,891],{"slug":746,"name":746,"fn":747,"description":748,"org":749,"tags":750,"stars":760,"repoUrl":761,"updatedAt":762},"nemoclaw-user-guide","retrieve NemoClaw documentation and configuration","Guides human users' AI agents to the NemoClaw docs MCP server and canonical Fern documentation in Markdown form. Use when users ask how to install, configure, operate, troubleshoot, secure, or learn NemoClaw with an AI coding assistant. Trigger keywords - nemoclaw docs, use nemoclaw with ai agent, nemoclaw mcp docs, nemoclaw install help, nemoclaw quickstart, nemoclaw markdown docs, llms.txt, agent skills.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[751,754,757],{"name":752,"slug":753,"type":15},"Documentation","documentation",{"name":755,"slug":756,"type":15},"MCP","mcp",{"name":758,"slug":759,"type":15},"Search","search",21777,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw","2026-07-20T06:00:01.461044",{"slug":764,"name":764,"fn":765,"description":766,"org":767,"tags":768,"stars":777,"repoUrl":778,"updatedAt":779},"mcore-build-and-dependency","manage Megatron-LM development environments","Container-based dev environment setup and dependency management for Megatron-LM. Covers acquiring and launching the CI container, uv package management, and updating uv.lock.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[769,772,775],{"name":770,"slug":771,"type":15},"Containers","containers",{"name":773,"slug":774,"type":15},"Deployment","deployment",{"name":776,"slug":95,"type":15},"Python",17049,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM","2026-07-27T06:06:11.249662",{"slug":781,"name":781,"fn":782,"description":783,"org":784,"tags":785,"stars":777,"repoUrl":778,"updatedAt":790},"mcore-bump-base-image","update NVIDIA PyTorch base images","Bump the NVIDIA PyTorch base image (`nvcr.io\u002Fnvidia\u002Fpytorch:YY.MM-py3`) used by Megatron-LM CI. Covers the two pin sites (GitHub CI in `docker\u002F.ngc_version.dev` and GitLab CI in `.gitlab\u002Fstages\u002F01.build.yml`), the post-bump CI loop (re-run functional tests, refresh golden values, mark broken tests), and the gotchas that bit PRs",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[786,789],{"name":787,"slug":788,"type":15},"CI\u002FCD","ci-cd",{"name":773,"slug":774,"type":15},"2026-07-14T05:25:59.97109",{"slug":792,"name":792,"fn":793,"description":794,"org":795,"tags":796,"stars":777,"repoUrl":778,"updatedAt":802},"mcore-cicd","manage CI\u002FCD pipelines for Megatron-LM","CI\u002FCD reference for Megatron-LM. Covers CI pipeline structure, PR scope labels, triggering internal GitLab CI (which force-pushes the current branch to a pull-request\u002FBRANCH ref — always dry-run and verify the destination first; never run against shared or protected branches), and CI failure investigation.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[797,798,799],{"name":787,"slug":788,"type":15},{"name":773,"slug":774,"type":15},{"name":800,"slug":801,"type":15},"GitHub","github","2026-07-27T06:06:12.278222",{"slug":804,"name":804,"fn":805,"description":806,"org":807,"tags":808,"stars":777,"repoUrl":778,"updatedAt":816},"mcore-create-issue","investigate CI failures and create issues","Investigate a failing GitHub Actions run or job and create a GitHub issue for the failure.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[809,812,813],{"name":810,"slug":811,"type":15},"Debugging","debugging",{"name":800,"slug":801,"type":15},{"name":814,"slug":815,"type":15},"Triage","triage","2026-07-14T05:25:57.442089",{"slug":818,"name":818,"fn":819,"description":820,"org":821,"tags":822,"stars":777,"repoUrl":778,"updatedAt":829},"mcore-linting-and-formatting","lint and format Megatron-LM code","Linting and formatting for Megatron-LM. Covers running autoformat.sh, tools (ruff, black, isort, pylint, mypy), and code style rules.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[823,826],{"name":824,"slug":825,"type":15},"Best Practices","best-practices",{"name":827,"slug":828,"type":15},"Code Analysis","code-analysis","2026-07-14T05:25:56.18433",{"slug":831,"name":831,"fn":832,"description":833,"org":834,"tags":835,"stars":777,"repoUrl":778,"updatedAt":843},"mcore-migrate-gpt-to-hybrid","migrate Megatron-LM models to HybridModel","Migration guide for moving Megatron Core GPTModel checkpoints, model providers, training commands, and layer mappings to HybridModel.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[836,839,842],{"name":837,"slug":838,"type":15},"Machine Learning","machine-learning",{"name":840,"slug":841,"type":15},"Migration","migration",{"name":9,"slug":8,"type":15},"2026-07-17T06:07:11.777011",{"slug":845,"name":845,"fn":846,"description":847,"org":848,"tags":849,"stars":777,"repoUrl":778,"updatedAt":856},"mcore-onboard-gb200-1node-tests","onboard functional tests for GB200","Onboard 1-node GitHub MR functional tests for GB200 from existing mr-scoped 2-node tests.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[850,853],{"name":851,"slug":852,"type":15},"QA","qa",{"name":854,"slug":855,"type":15},"Testing","testing","2026-07-14T05:25:53.673039",{"slug":858,"name":858,"fn":859,"description":860,"org":861,"tags":862,"stars":777,"repoUrl":778,"updatedAt":867},"mcore-run-on-slurm","launch distributed training jobs on SLURM","How to launch distributed Megatron-LM training jobs on a SLURM cluster. Covers a minimal sbatch skeleton, environment-variable setup for torch.distributed.run, CUDA_DEVICE_MAX_CONNECTIONS rules across hardware and parallelism modes, container conventions, monitoring, and per-rank failure diagnosis.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[863,864],{"name":773,"slug":774,"type":15},{"name":865,"slug":866,"type":15},"Infrastructure","infrastructure","2026-07-14T05:25:49.362534",{"slug":869,"name":869,"fn":870,"description":871,"org":872,"tags":873,"stars":777,"repoUrl":778,"updatedAt":881},"mcore-split-pr","split pull requests to reduce review load","Split a PR into multiple PRs to reduce the number of required CODEOWNERS reviewer groups.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[874,877,878],{"name":875,"slug":876,"type":15},"Code Review","code-review",{"name":800,"slug":801,"type":15},{"name":879,"slug":880,"type":15},"Pull Requests","pull-requests","2026-07-14T05:26:01.226578",{"slug":883,"name":883,"fn":884,"description":885,"org":886,"tags":887,"stars":777,"repoUrl":778,"updatedAt":890},"mcore-testing","run and manage Megatron-LM tests","Test system for Megatron-LM. Covers test layout, recipe YAML structure, adding and running unit and functional tests, golden values, marker filters, and CI parity.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[888,889],{"name":851,"slug":852,"type":15},{"name":854,"slug":855,"type":15},"2026-07-14T05:25:54.928983",{"slug":892,"name":892,"fn":893,"description":894,"org":895,"tags":896,"stars":777,"repoUrl":778,"updatedAt":901},"nightly-sync","manage nightly main-to-dev sync workflows","Domain knowledge for the nightly main-to-dev sync workflow. Covers merge strategy, CI architecture, failure investigation, and known issues.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[897,900],{"name":898,"slug":899,"type":15},"Automation","automation",{"name":787,"slug":788,"type":15},"2026-07-30T05:29:03.275638",496,{"items":904,"total":995},[905,917,930,945,960,967,982],{"slug":906,"name":906,"fn":907,"description":908,"org":909,"tags":910,"stars":23,"repoUrl":24,"updatedAt":916},"analysis-methods","write Python analysis code for FHIR data","Teaches the analyst agent how to write correct, robust Python analysis code for FHIR clinical data using pandas, matplotlib, and scipy.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[911,912,913,914,915],{"name":17,"slug":18,"type":15},{"name":21,"slug":22,"type":15},{"name":13,"slug":14,"type":15},{"name":9,"slug":8,"type":15},{"name":776,"slug":95,"type":15},"2026-07-14T05:35:59.037962",{"slug":918,"name":918,"fn":919,"description":920,"org":921,"tags":922,"stars":23,"repoUrl":24,"updatedAt":929},"case-summary","summarize clinical patient cases from FHIR","Prepare a complete clinical case summary for a patient from FHIR endpoints. Use when asked to summarize a patient, compile a case, or prepare for tumor board.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[923,924,925,926],{"name":21,"slug":22,"type":15},{"name":13,"slug":14,"type":15},{"name":9,"slug":8,"type":15},{"name":927,"slug":928,"type":15},"Summarization","summarization","2026-07-14T05:35:52.790528",{"slug":931,"name":931,"fn":932,"description":933,"org":934,"tags":935,"stars":23,"repoUrl":24,"updatedAt":944},"clinical-delegation","delegate clinical tasks to specialist agents","How to delegate clinical tasks to specialist agents. Always use sub-agent runtime with explicit agentId — never ACP. Never call FHIR via web_fetch.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[936,939,940,943],{"name":937,"slug":938,"type":15},"Agents","agents",{"name":13,"slug":14,"type":15},{"name":941,"slug":942,"type":15},"Multi-Agent","multi-agent",{"name":9,"slug":8,"type":15},"2026-07-14T05:35:55.294972",{"slug":946,"name":946,"fn":947,"description":948,"org":949,"tags":950,"stars":23,"repoUrl":24,"updatedAt":959},"clinical-knowledge","provide clinical reference and regulatory context","Teaches agents clinical reference ranges, condition codes, quality measure definitions, drug classifications, and regulatory context so they can flag abnormal values and identify care gaps.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[951,954,955,956],{"name":952,"slug":953,"type":15},"Clinical Trials","clinical-trials",{"name":13,"slug":14,"type":15},{"name":9,"slug":8,"type":15},{"name":957,"slug":958,"type":15},"Regulatory Compliance","regulatory-compliance","2026-07-14T05:35:56.550833",{"slug":4,"name":4,"fn":5,"description":6,"org":961,"tags":962,"stars":23,"repoUrl":24,"updatedAt":25},{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[963,964,965,966],{"name":17,"slug":18,"type":15},{"name":21,"slug":22,"type":15},{"name":13,"slug":14,"type":15},{"name":9,"slug":8,"type":15},{"slug":968,"name":968,"fn":969,"description":970,"org":971,"tags":972,"stars":23,"repoUrl":24,"updatedAt":981},"dgx-diagnose","diagnose NVIDIA DGX Station hardware issues","Diagnose common DGX Station GB300 issues — CUDA crashes, wrong-GPU targeting, vLLM\u002FSGLang container bugs, MIG state problems, NVLink\u002FFabric Manager errors, X\u002FVulkan failures, HuggingFace auth, and port conflicts. Use when the user reports a GPU error, inference server crash, MIG problem, or any unexplained DGX Station failure.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[973,976,977,978],{"name":974,"slug":975,"type":15},"AI Infrastructure","ai-infrastructure",{"name":810,"slug":811,"type":15},{"name":9,"slug":8,"type":15},{"name":979,"slug":980,"type":15},"Observability","observability","2026-07-14T05:31:04.085598",{"slug":983,"name":983,"fn":984,"description":985,"org":986,"tags":987,"stars":23,"repoUrl":24,"updatedAt":994},"fhir-basics","query and parse FHIR R4 API resources","Teaches agents how FHIR R4 APIs work, what resources are available, how to query them with search parameters, and how to correctly parse all response formats including component Observations.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[988,991,992,993],{"name":989,"slug":990,"type":15},"API Development","api-development",{"name":21,"slug":22,"type":15},{"name":13,"slug":14,"type":15},{"name":9,"slug":8,"type":15},"2026-07-14T05:35:57.797226",11]