[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-anthropic-clinical-note-extract-skill":3,"mdc--68yrqe-key":30,"related-repo-anthropic-clinical-note-extract-skill":902,"related-org-anthropic-clinical-note-extract-skill":1006},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":12,"stars":20,"repoUrl":21,"updatedAt":22,"license":23,"forks":24,"topics":25,"repo":26,"sourceUrl":28,"mdContent":29},"clinical-note-extract-skill","extract structured data from clinical notes","Extract structured data from clinical notes with span-level provenance and null-safety. Use when users say \"extract [variables] from this note\", \"abstract this chart\", \"pull structured data from these notes\", \"what does this note say about [field]\", or when building a chart-abstraction, registry, or cohort dataset from unstructured clinical text.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},"anthropic","Anthropic","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fanthropic.png","anthropics",[13,17],{"name":14,"slug":15,"type":16},"Healthcare","healthcare","tag",{"name":18,"slug":19,"type":16},"Data Analysis","data-analysis",356,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fhealthcare","2026-06-28T07:57:18.332447",null,91,[],{"repoUrl":21,"stars":20,"forks":24,"topics":27,"description":23},[],"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fhealthcare\u002Ftree\u002FHEAD\u002Fplugins\u002Fhealthcare\u002Fskills\u002Fclinical-note-extract","---\nname: clinical-note-extract-skill\ndescription: Extract structured data from clinical notes with span-level provenance and null-safety. Use when users say \"extract [variables] from this note\", \"abstract this chart\", \"pull structured data from these notes\", \"what does this note say about [field]\", or when building a chart-abstraction, registry, or cohort dataset from unstructured clinical text.\n---\n\n# Clinical Note Extraction\n\nStructured extraction from clinical notes against a user-defined schema, with span citations for every value and explicit nulls for every absence. One note or many — the path is the same: an isolated no-tools worker extracts each note, then a deterministic validation pass verifies spans and codes.\n\nThis is the extraction primitive that care-gap reasoning, adverse-event detection, trial-eligibility screening, prior-auth evidence assembly, and registry abstraction sit on.\n\n## Steps\n\n```\n1  Define schema   — references\u002F01-define-schema.md\n2  Extract         — workflows\u002Fextract-batch.js (one isolated worker per note)\n3  Validate        — span check + run each field's `check`\n4  Report          — references\u002F03-review.md\n```\n\n### Step 1 — Define schema\n\nRead `references\u002F01-define-schema.md`. Turn the user's request into a schema: each field is `{desc, finding?, check?}`. `desc` says what to look for in the note's own terms; `finding: true` means classify assertion; `check` is how step 3 validates (open-ended — `{kind: \"terminology\"|\"range\"|\"date\"|\"pattern\"|\"enum\"|..., ...params}`). Confirm with the user before extracting.\n\n### Step 2 — Extract\n\nHowever the user supplied notes — pasted text, file paths, a directory, PDFs, a FHIR connector, a database query — resolve each to plain text using whatever tools you have, then call the saved workflow with one `{id, text}` per note. The workflow's input contract is the only strict piece; how you get there is yours to figure out. It runs one `note-extract-worker` agent per note (no tools — note text is untrusted), each following `references\u002Frules.md`, and returns one schema-enforced record per note:\n\n```\nWorkflow({\n  scriptPath: \"\u003Cthis skill dir>\u002Fworkflows\u002Fextract-batch.js\",\n  args: {\n    notes:  [{id, text}, ...],     \u002F\u002F one or many\n    schema: \u003Cthe schema from step 1>,\n    rules:  \u003CRead references\u002Frules.md verbatim>\n  }\n})\n```\n\nWorkers have no tools — they return only what they read (`value`, `span`, `presence`\u002F`temporality`\u002F`experiencer`, `null_reason`, `unit`). All checks happen in step 3. Because note text rides inline in `args`, the workflow path tops out at a few dozen notes per call. For larger corpora, run `bun \u003Cthis skill dir>\u002Fscripts\u002Fbatch.ts \u003Cnotes-dir> \u003Cschema.json> records.jsonl` instead — it reads files in trusted code and spawns one tool-disabled extraction per note with the same rules, then resume at step 3 over the resulting `records.jsonl`.\n\n### Step 3 — Validate\n\nRuns here in the calling session. Deterministic — no model judgment. For every record:\n\n1. **Span check.** For every non-null field, confirm `span` appears verbatim in that note's source text. Attach `span_verified`.\n2. **Run each field's `check`.** Dispatch on `check.kind`:\n   - `terminology` — dedupe `(check.via, value)` across all records, look each up via whatever connector answers to `via`, attach `{code, code_status, display}`. No connector for that `via` → `code_status: \"unvalidated\"`, name it in the report.\n   - `range` — `value` vs `[min, max]` and `unit` vs `check.unit`; attach `range_flag`.\n   - `date` — confirm `value` parses as a date; attach `date_ok`.\n   - `pattern` \u002F `enum` — match; attach `check_ok`.\n   - other \u002F no `check` — nothing to attach.\n\nA field is trustworthy when `span_verified` and its check (if any) passed. Adding a check kind = add a branch here; nothing upstream changes.\n\n### Step 4 — Report\n\nRead `references\u002F03-review.md`. Produce one row per (note, field): `note_id | field | value | presence\u002Ftemporality\u002Fexperiencer | span | check`. Below it, the completion summary: fields requested \u002F populated \u002F null, and per `check.kind` what passed vs flagged (name any terminology `via` that lacked a connector). Never let a failed check or unverified span pass silently.\n\nOffer to write records + report to `~\u002F.claude\u002Fdata\u002Fhealthcare\u002Fclinical-note-extract\u002F\u003Crun-id>\u002F`. That directory is local working state, not an archive: do not copy it to shared drives or external systems without the user's explicit instruction, and tell the user it can be deleted once they have what they need — extracted records carry whatever PHI was in the source notes.\n\n## Output contract\n\nWorker emits, per field: `{value, span, location, presence?, temporality?, experiencer?, null_reason?, unit?}` — only what it read. Step 3 attaches `span_verified` plus whatever the field's `check` produced (`code`\u002F`code_status`\u002F`display` for terminology, `range_flag` for range, etc.).\n\n### Optional — export as FHIR\n\nIf the user wants FHIR resources instead of flat records, the assertion axes map directly:\n\n| record | FHIR |\n|---|---|\n| `experiencer != patient` | `FamilyMemberHistory.condition` (not `Condition`) |\n| `presence: absent` → `verificationStatus: refuted`; `possible` → `unconfirmed`; `present` → `confirmed` | `Condition.verificationStatus` |\n| `temporality: historical` → `inactive`; `current` → `active` | `Condition.clinicalStatus` |\n| `temporality: hypothetical` | no native field — omit, or use a `RiskAssessment` resource |\n| `value` + terminology check result | `Condition.code` as a `CodeableConcept` (`{text: value, coding: [{system, code, display}]}`) |\n| `span` + `location` | `Condition.note` or a provenance extension |\n\nThis is a deterministic transform over the validated records — no model call. Offer it when the user names FHIR as the target; otherwise the flat records are the default.\n\n## Prerequisites\n\nConnectors for whatever `check.via` values the schema names. Missing ones don't block extraction — those fields stay unvalidated and the report names them.\n",{"data":31,"body":32},{"name":4,"description":6},{"type":33,"children":34},"root",[35,44,50,55,62,75,82,135,141,170,179,260,266,271,489,501,507,541,554,560,615,621,626,878,883,889],{"type":36,"tag":37,"props":38,"children":40},"element","h1",{"id":39},"clinical-note-extraction",[41],{"type":42,"value":43},"text","Clinical Note Extraction",{"type":36,"tag":45,"props":46,"children":47},"p",{},[48],{"type":42,"value":49},"Structured extraction from clinical notes against a user-defined schema, with span citations for every value and explicit nulls for every absence. One note or many — the path is the same: an isolated no-tools worker extracts each note, then a deterministic validation pass verifies spans and codes.",{"type":36,"tag":45,"props":51,"children":52},{},[53],{"type":42,"value":54},"This is the extraction primitive that care-gap reasoning, adverse-event detection, trial-eligibility screening, prior-auth evidence assembly, and registry abstraction sit on.",{"type":36,"tag":56,"props":57,"children":59},"h2",{"id":58},"steps",[60],{"type":42,"value":61},"Steps",{"type":36,"tag":63,"props":64,"children":68},"pre",{"className":65,"code":67,"language":42},[66],"language-text","1  Define schema   — references\u002F01-define-schema.md\n2  Extract         — workflows\u002Fextract-batch.js (one isolated worker per note)\n3  Validate        — span check + run each field's `check`\n4  Report          — references\u002F03-review.md\n",[69],{"type":36,"tag":70,"props":71,"children":73},"code",{"__ignoreMap":72},"",[74],{"type":42,"value":67},{"type":36,"tag":76,"props":77,"children":79},"h3",{"id":78},"step-1-define-schema",[80],{"type":42,"value":81},"Step 1 — Define schema",{"type":36,"tag":45,"props":83,"children":84},{},[85,87,93,95,101,103,109,111,117,119,125,127,133],{"type":42,"value":86},"Read ",{"type":36,"tag":70,"props":88,"children":90},{"className":89},[],[91],{"type":42,"value":92},"references\u002F01-define-schema.md",{"type":42,"value":94},". Turn the user's request into a schema: each field is ",{"type":36,"tag":70,"props":96,"children":98},{"className":97},[],[99],{"type":42,"value":100},"{desc, finding?, check?}",{"type":42,"value":102},". ",{"type":36,"tag":70,"props":104,"children":106},{"className":105},[],[107],{"type":42,"value":108},"desc",{"type":42,"value":110}," says what to look for in the note's own terms; ",{"type":36,"tag":70,"props":112,"children":114},{"className":113},[],[115],{"type":42,"value":116},"finding: true",{"type":42,"value":118}," means classify assertion; ",{"type":36,"tag":70,"props":120,"children":122},{"className":121},[],[123],{"type":42,"value":124},"check",{"type":42,"value":126}," is how step 3 validates (open-ended — ",{"type":36,"tag":70,"props":128,"children":130},{"className":129},[],[131],{"type":42,"value":132},"{kind: \"terminology\"|\"range\"|\"date\"|\"pattern\"|\"enum\"|..., ...params}",{"type":42,"value":134},"). Confirm with the user before extracting.",{"type":36,"tag":76,"props":136,"children":138},{"id":137},"step-2-extract",[139],{"type":42,"value":140},"Step 2 — Extract",{"type":36,"tag":45,"props":142,"children":143},{},[144,146,152,154,160,162,168],{"type":42,"value":145},"However the user supplied notes — pasted text, file paths, a directory, PDFs, a FHIR connector, a database query — resolve each to plain text using whatever tools you have, then call the saved workflow with one ",{"type":36,"tag":70,"props":147,"children":149},{"className":148},[],[150],{"type":42,"value":151},"{id, text}",{"type":42,"value":153}," per note. The workflow's input contract is the only strict piece; how you get there is yours to figure out. It runs one ",{"type":36,"tag":70,"props":155,"children":157},{"className":156},[],[158],{"type":42,"value":159},"note-extract-worker",{"type":42,"value":161}," agent per note (no tools — note text is untrusted), each following ",{"type":36,"tag":70,"props":163,"children":165},{"className":164},[],[166],{"type":42,"value":167},"references\u002Frules.md",{"type":42,"value":169},", and returns one schema-enforced record per note:",{"type":36,"tag":63,"props":171,"children":174},{"className":172,"code":173,"language":42},[66],"Workflow({\n  scriptPath: \"\u003Cthis skill dir>\u002Fworkflows\u002Fextract-batch.js\",\n  args: {\n    notes:  [{id, text}, ...],     \u002F\u002F one or many\n    schema: \u003Cthe schema from step 1>,\n    rules:  \u003CRead references\u002Frules.md verbatim>\n  }\n})\n",[175],{"type":36,"tag":70,"props":176,"children":177},{"__ignoreMap":72},[178],{"type":42,"value":173},{"type":36,"tag":45,"props":180,"children":181},{},[182,184,190,192,198,199,205,207,213,214,220,221,227,228,234,236,242,244,250,252,258],{"type":42,"value":183},"Workers have no tools — they return only what they read (",{"type":36,"tag":70,"props":185,"children":187},{"className":186},[],[188],{"type":42,"value":189},"value",{"type":42,"value":191},", ",{"type":36,"tag":70,"props":193,"children":195},{"className":194},[],[196],{"type":42,"value":197},"span",{"type":42,"value":191},{"type":36,"tag":70,"props":200,"children":202},{"className":201},[],[203],{"type":42,"value":204},"presence",{"type":42,"value":206},"\u002F",{"type":36,"tag":70,"props":208,"children":210},{"className":209},[],[211],{"type":42,"value":212},"temporality",{"type":42,"value":206},{"type":36,"tag":70,"props":215,"children":217},{"className":216},[],[218],{"type":42,"value":219},"experiencer",{"type":42,"value":191},{"type":36,"tag":70,"props":222,"children":224},{"className":223},[],[225],{"type":42,"value":226},"null_reason",{"type":42,"value":191},{"type":36,"tag":70,"props":229,"children":231},{"className":230},[],[232],{"type":42,"value":233},"unit",{"type":42,"value":235},"). All checks happen in step 3. Because note text rides inline in ",{"type":36,"tag":70,"props":237,"children":239},{"className":238},[],[240],{"type":42,"value":241},"args",{"type":42,"value":243},", the workflow path tops out at a few dozen notes per call. For larger corpora, run ",{"type":36,"tag":70,"props":245,"children":247},{"className":246},[],[248],{"type":42,"value":249},"bun \u003Cthis skill dir>\u002Fscripts\u002Fbatch.ts \u003Cnotes-dir> \u003Cschema.json> records.jsonl",{"type":42,"value":251}," instead — it reads files in trusted code and spawns one tool-disabled extraction per note with the same rules, then resume at step 3 over the resulting ",{"type":36,"tag":70,"props":253,"children":255},{"className":254},[],[256],{"type":42,"value":257},"records.jsonl",{"type":42,"value":259},".",{"type":36,"tag":76,"props":261,"children":263},{"id":262},"step-3-validate",[264],{"type":42,"value":265},"Step 3 — Validate",{"type":36,"tag":45,"props":267,"children":268},{},[269],{"type":42,"value":270},"Runs here in the calling session. Deterministic — no model judgment. For every record:",{"type":36,"tag":272,"props":273,"children":274},"ol",{},[275,301],{"type":36,"tag":276,"props":277,"children":278},"li",{},[279,285,287,292,294,300],{"type":36,"tag":280,"props":281,"children":282},"strong",{},[283],{"type":42,"value":284},"Span check.",{"type":42,"value":286}," For every non-null field, confirm ",{"type":36,"tag":70,"props":288,"children":290},{"className":289},[],[291],{"type":42,"value":197},{"type":42,"value":293}," appears verbatim in that note's source text. Attach ",{"type":36,"tag":70,"props":295,"children":297},{"className":296},[],[298],{"type":42,"value":299},"span_verified",{"type":42,"value":259},{"type":36,"tag":276,"props":302,"children":303},{},[304,315,317,323,325],{"type":36,"tag":280,"props":305,"children":306},{},[307,309,314],{"type":42,"value":308},"Run each field's ",{"type":36,"tag":70,"props":310,"children":312},{"className":311},[],[313],{"type":42,"value":124},{"type":42,"value":259},{"type":42,"value":316}," Dispatch on ",{"type":36,"tag":70,"props":318,"children":320},{"className":319},[],[321],{"type":42,"value":322},"check.kind",{"type":42,"value":324},":\n",{"type":36,"tag":326,"props":327,"children":328},"ul",{},[329,379,426,451,477],{"type":36,"tag":276,"props":330,"children":331},{},[332,338,340,346,348,354,356,362,364,369,371,377],{"type":36,"tag":70,"props":333,"children":335},{"className":334},[],[336],{"type":42,"value":337},"terminology",{"type":42,"value":339}," — dedupe ",{"type":36,"tag":70,"props":341,"children":343},{"className":342},[],[344],{"type":42,"value":345},"(check.via, value)",{"type":42,"value":347}," across all records, look each up via whatever connector answers to ",{"type":36,"tag":70,"props":349,"children":351},{"className":350},[],[352],{"type":42,"value":353},"via",{"type":42,"value":355},", attach ",{"type":36,"tag":70,"props":357,"children":359},{"className":358},[],[360],{"type":42,"value":361},"{code, code_status, display}",{"type":42,"value":363},". No connector for that ",{"type":36,"tag":70,"props":365,"children":367},{"className":366},[],[368],{"type":42,"value":353},{"type":42,"value":370}," → ",{"type":36,"tag":70,"props":372,"children":374},{"className":373},[],[375],{"type":42,"value":376},"code_status: \"unvalidated\"",{"type":42,"value":378},", name it in the report.",{"type":36,"tag":276,"props":380,"children":381},{},[382,388,390,395,397,403,405,410,411,417,419,425],{"type":36,"tag":70,"props":383,"children":385},{"className":384},[],[386],{"type":42,"value":387},"range",{"type":42,"value":389}," — ",{"type":36,"tag":70,"props":391,"children":393},{"className":392},[],[394],{"type":42,"value":189},{"type":42,"value":396}," vs ",{"type":36,"tag":70,"props":398,"children":400},{"className":399},[],[401],{"type":42,"value":402},"[min, max]",{"type":42,"value":404}," and ",{"type":36,"tag":70,"props":406,"children":408},{"className":407},[],[409],{"type":42,"value":233},{"type":42,"value":396},{"type":36,"tag":70,"props":412,"children":414},{"className":413},[],[415],{"type":42,"value":416},"check.unit",{"type":42,"value":418},"; attach ",{"type":36,"tag":70,"props":420,"children":422},{"className":421},[],[423],{"type":42,"value":424},"range_flag",{"type":42,"value":259},{"type":36,"tag":276,"props":427,"children":428},{},[429,435,437,442,444,450],{"type":36,"tag":70,"props":430,"children":432},{"className":431},[],[433],{"type":42,"value":434},"date",{"type":42,"value":436}," — confirm ",{"type":36,"tag":70,"props":438,"children":440},{"className":439},[],[441],{"type":42,"value":189},{"type":42,"value":443}," parses as a date; attach ",{"type":36,"tag":70,"props":445,"children":447},{"className":446},[],[448],{"type":42,"value":449},"date_ok",{"type":42,"value":259},{"type":36,"tag":276,"props":452,"children":453},{},[454,460,462,468,470,476],{"type":36,"tag":70,"props":455,"children":457},{"className":456},[],[458],{"type":42,"value":459},"pattern",{"type":42,"value":461}," \u002F ",{"type":36,"tag":70,"props":463,"children":465},{"className":464},[],[466],{"type":42,"value":467},"enum",{"type":42,"value":469}," — match; attach ",{"type":36,"tag":70,"props":471,"children":473},{"className":472},[],[474],{"type":42,"value":475},"check_ok",{"type":42,"value":259},{"type":36,"tag":276,"props":478,"children":479},{},[480,482,487],{"type":42,"value":481},"other \u002F no ",{"type":36,"tag":70,"props":483,"children":485},{"className":484},[],[486],{"type":42,"value":124},{"type":42,"value":488}," — nothing to attach.",{"type":36,"tag":45,"props":490,"children":491},{},[492,494,499],{"type":42,"value":493},"A field is trustworthy when ",{"type":36,"tag":70,"props":495,"children":497},{"className":496},[],[498],{"type":42,"value":299},{"type":42,"value":500}," and its check (if any) passed. Adding a check kind = add a branch here; nothing upstream changes.",{"type":36,"tag":76,"props":502,"children":504},{"id":503},"step-4-report",[505],{"type":42,"value":506},"Step 4 — Report",{"type":36,"tag":45,"props":508,"children":509},{},[510,511,517,519,525,527,532,534,539],{"type":42,"value":86},{"type":36,"tag":70,"props":512,"children":514},{"className":513},[],[515],{"type":42,"value":516},"references\u002F03-review.md",{"type":42,"value":518},". Produce one row per (note, field): ",{"type":36,"tag":70,"props":520,"children":522},{"className":521},[],[523],{"type":42,"value":524},"note_id | field | value | presence\u002Ftemporality\u002Fexperiencer | span | check",{"type":42,"value":526},". Below it, the completion summary: fields requested \u002F populated \u002F null, and per ",{"type":36,"tag":70,"props":528,"children":530},{"className":529},[],[531],{"type":42,"value":322},{"type":42,"value":533}," what passed vs flagged (name any terminology ",{"type":36,"tag":70,"props":535,"children":537},{"className":536},[],[538],{"type":42,"value":353},{"type":42,"value":540}," that lacked a connector). Never let a failed check or unverified span pass silently.",{"type":36,"tag":45,"props":542,"children":543},{},[544,546,552],{"type":42,"value":545},"Offer to write records + report to ",{"type":36,"tag":70,"props":547,"children":549},{"className":548},[],[550],{"type":42,"value":551},"~\u002F.claude\u002Fdata\u002Fhealthcare\u002Fclinical-note-extract\u002F\u003Crun-id>\u002F",{"type":42,"value":553},". That directory is local working state, not an archive: do not copy it to shared drives or external systems without the user's explicit instruction, and tell the user it can be deleted once they have what they need — extracted records carry whatever PHI was in the source notes.",{"type":36,"tag":56,"props":555,"children":557},{"id":556},"output-contract",[558],{"type":42,"value":559},"Output contract",{"type":36,"tag":45,"props":561,"children":562},{},[563,565,571,573,578,580,585,587,592,593,599,600,606,608,613],{"type":42,"value":564},"Worker emits, per field: ",{"type":36,"tag":70,"props":566,"children":568},{"className":567},[],[569],{"type":42,"value":570},"{value, span, location, presence?, temporality?, experiencer?, null_reason?, unit?}",{"type":42,"value":572}," — only what it read. Step 3 attaches ",{"type":36,"tag":70,"props":574,"children":576},{"className":575},[],[577],{"type":42,"value":299},{"type":42,"value":579}," plus whatever the field's ",{"type":36,"tag":70,"props":581,"children":583},{"className":582},[],[584],{"type":42,"value":124},{"type":42,"value":586}," produced (",{"type":36,"tag":70,"props":588,"children":590},{"className":589},[],[591],{"type":42,"value":70},{"type":42,"value":206},{"type":36,"tag":70,"props":594,"children":596},{"className":595},[],[597],{"type":42,"value":598},"code_status",{"type":42,"value":206},{"type":36,"tag":70,"props":601,"children":603},{"className":602},[],[604],{"type":42,"value":605},"display",{"type":42,"value":607}," for terminology, ",{"type":36,"tag":70,"props":609,"children":611},{"className":610},[],[612],{"type":42,"value":424},{"type":42,"value":614}," for range, etc.).",{"type":36,"tag":76,"props":616,"children":618},{"id":617},"optional-export-as-fhir",[619],{"type":42,"value":620},"Optional — export as FHIR",{"type":36,"tag":45,"props":622,"children":623},{},[624],{"type":42,"value":625},"If the user wants FHIR resources instead of flat records, the assertion axes map directly:",{"type":36,"tag":627,"props":628,"children":629},"table",{},[630,649],{"type":36,"tag":631,"props":632,"children":633},"thead",{},[634],{"type":36,"tag":635,"props":636,"children":637},"tr",{},[638,644],{"type":36,"tag":639,"props":640,"children":641},"th",{},[642],{"type":42,"value":643},"record",{"type":36,"tag":639,"props":645,"children":646},{},[647],{"type":42,"value":648},"FHIR",{"type":36,"tag":650,"props":651,"children":652},"tbody",{},[653,685,742,784,809,848],{"type":36,"tag":635,"props":654,"children":655},{},[656,666],{"type":36,"tag":657,"props":658,"children":659},"td",{},[660],{"type":36,"tag":70,"props":661,"children":663},{"className":662},[],[664],{"type":42,"value":665},"experiencer != patient",{"type":36,"tag":657,"props":667,"children":668},{},[669,675,677,683],{"type":36,"tag":70,"props":670,"children":672},{"className":671},[],[673],{"type":42,"value":674},"FamilyMemberHistory.condition",{"type":42,"value":676}," (not ",{"type":36,"tag":70,"props":678,"children":680},{"className":679},[],[681],{"type":42,"value":682},"Condition",{"type":42,"value":684},")",{"type":36,"tag":635,"props":686,"children":687},{},[688,733],{"type":36,"tag":657,"props":689,"children":690},{},[691,697,698,704,706,712,713,719,720,726,727],{"type":36,"tag":70,"props":692,"children":694},{"className":693},[],[695],{"type":42,"value":696},"presence: absent",{"type":42,"value":370},{"type":36,"tag":70,"props":699,"children":701},{"className":700},[],[702],{"type":42,"value":703},"verificationStatus: refuted",{"type":42,"value":705},"; ",{"type":36,"tag":70,"props":707,"children":709},{"className":708},[],[710],{"type":42,"value":711},"possible",{"type":42,"value":370},{"type":36,"tag":70,"props":714,"children":716},{"className":715},[],[717],{"type":42,"value":718},"unconfirmed",{"type":42,"value":705},{"type":36,"tag":70,"props":721,"children":723},{"className":722},[],[724],{"type":42,"value":725},"present",{"type":42,"value":370},{"type":36,"tag":70,"props":728,"children":730},{"className":729},[],[731],{"type":42,"value":732},"confirmed",{"type":36,"tag":657,"props":734,"children":735},{},[736],{"type":36,"tag":70,"props":737,"children":739},{"className":738},[],[740],{"type":42,"value":741},"Condition.verificationStatus",{"type":36,"tag":635,"props":743,"children":744},{},[745,775],{"type":36,"tag":657,"props":746,"children":747},{},[748,754,755,761,762,768,769],{"type":36,"tag":70,"props":749,"children":751},{"className":750},[],[752],{"type":42,"value":753},"temporality: historical",{"type":42,"value":370},{"type":36,"tag":70,"props":756,"children":758},{"className":757},[],[759],{"type":42,"value":760},"inactive",{"type":42,"value":705},{"type":36,"tag":70,"props":763,"children":765},{"className":764},[],[766],{"type":42,"value":767},"current",{"type":42,"value":370},{"type":36,"tag":70,"props":770,"children":772},{"className":771},[],[773],{"type":42,"value":774},"active",{"type":36,"tag":657,"props":776,"children":777},{},[778],{"type":36,"tag":70,"props":779,"children":781},{"className":780},[],[782],{"type":42,"value":783},"Condition.clinicalStatus",{"type":36,"tag":635,"props":785,"children":786},{},[787,796],{"type":36,"tag":657,"props":788,"children":789},{},[790],{"type":36,"tag":70,"props":791,"children":793},{"className":792},[],[794],{"type":42,"value":795},"temporality: hypothetical",{"type":36,"tag":657,"props":797,"children":798},{},[799,801,807],{"type":42,"value":800},"no native field — omit, or use a ",{"type":36,"tag":70,"props":802,"children":804},{"className":803},[],[805],{"type":42,"value":806},"RiskAssessment",{"type":42,"value":808}," resource",{"type":36,"tag":635,"props":810,"children":811},{},[812,822],{"type":36,"tag":657,"props":813,"children":814},{},[815,820],{"type":36,"tag":70,"props":816,"children":818},{"className":817},[],[819],{"type":42,"value":189},{"type":42,"value":821}," + terminology check result",{"type":36,"tag":657,"props":823,"children":824},{},[825,831,833,839,841,847],{"type":36,"tag":70,"props":826,"children":828},{"className":827},[],[829],{"type":42,"value":830},"Condition.code",{"type":42,"value":832}," as a ",{"type":36,"tag":70,"props":834,"children":836},{"className":835},[],[837],{"type":42,"value":838},"CodeableConcept",{"type":42,"value":840}," (",{"type":36,"tag":70,"props":842,"children":844},{"className":843},[],[845],{"type":42,"value":846},"{text: value, coding: [{system, code, display}]}",{"type":42,"value":684},{"type":36,"tag":635,"props":849,"children":850},{},[851,867],{"type":36,"tag":657,"props":852,"children":853},{},[854,859,861],{"type":36,"tag":70,"props":855,"children":857},{"className":856},[],[858],{"type":42,"value":197},{"type":42,"value":860}," + ",{"type":36,"tag":70,"props":862,"children":864},{"className":863},[],[865],{"type":42,"value":866},"location",{"type":36,"tag":657,"props":868,"children":869},{},[870,876],{"type":36,"tag":70,"props":871,"children":873},{"className":872},[],[874],{"type":42,"value":875},"Condition.note",{"type":42,"value":877}," or a provenance extension",{"type":36,"tag":45,"props":879,"children":880},{},[881],{"type":42,"value":882},"This is a deterministic transform over the validated records — no model call. Offer it when the user names FHIR as the target; otherwise the flat records are the default.",{"type":36,"tag":56,"props":884,"children":886},{"id":885},"prerequisites",[887],{"type":42,"value":888},"Prerequisites",{"type":36,"tag":45,"props":890,"children":891},{},[892,894,900],{"type":42,"value":893},"Connectors for whatever ",{"type":36,"tag":70,"props":895,"children":897},{"className":896},[],[898],{"type":42,"value":899},"check.via",{"type":42,"value":901}," values the schema names. Missing ones don't block extraction — those fields stay unvalidated and the report names them.",{"items":903,"total":1005},[904,909,927,947,962,973,990],{"slug":4,"name":4,"fn":5,"description":6,"org":905,"tags":906,"stars":20,"repoUrl":21,"updatedAt":22},{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[907,908],{"name":18,"slug":19,"type":16},{"name":14,"slug":15,"type":16},{"slug":910,"name":910,"fn":911,"description":912,"org":913,"tags":914,"stars":20,"repoUrl":21,"updatedAt":926},"contracts","analyze contract documents with citations","Answer a question across a corpus of contract documents with verified citations. Use when the user asks what a contract says, which contracts have a clause, what changed between amendments, or any question that needs reading and citing across a set of contract files. The corpus must be on the local filesystem (see README).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[915,917,920,923],{"name":916,"slug":910,"type":16},"Contracts",{"name":918,"slug":919,"type":16},"Documents","documents",{"name":921,"slug":922,"type":16},"Legal","legal",{"name":924,"slug":925,"type":16},"Research","research","2026-07-18T05:15:17.073689",{"slug":928,"name":928,"fn":929,"description":930,"org":931,"tags":932,"stars":20,"repoUrl":21,"updatedAt":946},"doc-extract","extract text from documents","Extract plain text from a document file - PDF, DOCX, XLSX, PPTX, RTF, or plain text\u002Fmarkdown\u002FHTML. Use when a binary document needs to be turned into text, for example a contract PDF or an EHR DocumentReference attachment. Other skills (fhir) invoke scripts\u002Fextract.ts directly; the contracts MCP server bundles its own copy (servers\u002Fdocuments\u002Fsrc\u002Fextract.mjs) so its bundle stays self-contained — port fixes to both.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[933,936,937,940,943],{"name":934,"slug":935,"type":16},"Data Cleaning","data-cleaning",{"name":918,"slug":919,"type":16},{"name":938,"slug":939,"type":16},"DOCX","docx",{"name":941,"slug":942,"type":16},"PDF","pdf",{"name":944,"slug":945,"type":16},"Spreadsheets","spreadsheets","2026-07-18T05:15:15.766116",{"slug":948,"name":948,"fn":949,"description":950,"org":951,"tags":952,"stars":20,"repoUrl":21,"updatedAt":961},"fhir","extract clinical data from FHIR servers","Connect to a hospital's FHIR R4 server (Epic, Oracle Health\u002FCerner, MEDITECH, athenahealth, or any SMART-on-FHIR endpoint), pull a patient's clinical data and notes, and extract structured findings. Use when users say \"connect to the EHR\", \"connect to Epic\u002FCerner\", \"pull notes for patient X\", \"what do the last 6 months of notes say about Y\", or any task that starts from a live EHR rather than pasted text.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[953,956,957,958],{"name":954,"slug":955,"type":16},"API Development","api-development",{"name":648,"slug":948,"type":16},{"name":14,"slug":15,"type":16},{"name":959,"slug":960,"type":16},"Interoperability","interoperability","2026-07-03T16:28:01.476883",{"slug":963,"name":963,"fn":964,"description":965,"org":966,"tags":967,"stars":20,"repoUrl":21,"updatedAt":972},"fhir-developer-skill","build FHIR REST healthcare endpoints","FHIR API development guide for building healthcare endpoints. Use when: (1) Creating FHIR REST endpoints (Patient, Observation, Encounter, Condition, MedicationRequest), (2) Validating FHIR resources and returning proper HTTP status codes and error responses, (3) Implementing SMART on FHIR authorization and OAuth scopes, (4) Working with Bundles, transactions, batch operations, or search pagination. Covers FHIR R4 resource structures, required fields, value sets (status codes, gender, intent), coding systems (LOINC, SNOMED, RxNorm, ICD-10), and OperationOutcome error handling.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[968,969,970,971],{"name":954,"slug":955,"type":16},{"name":648,"slug":948,"type":16},{"name":14,"slug":15,"type":16},{"name":959,"slug":960,"type":16},"2026-06-16T09:39:06.471928",{"slug":974,"name":974,"fn":975,"description":976,"org":977,"tags":978,"stars":20,"repoUrl":21,"updatedAt":989},"fraud-detection","detect fraud in healthcare claims","Screen a Medicare\u002FMedicaid claims corpus for fraud, waste, and abuse and produce ranked, fully-cited investigation referrals for an SIU \u002F program-integrity team. Use when asked to run a fraud sweep, screen claims for FWA, find billing anomalies, or generate investigation referrals over a claims dataset.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[979,982,985,986],{"name":980,"slug":981,"type":16},"Audit","audit",{"name":983,"slug":984,"type":16},"Compliance","compliance",{"name":14,"slug":15,"type":16},{"name":987,"slug":988,"type":16},"Insurance","insurance","2026-06-26T07:50:49.129616",{"slug":991,"name":991,"fn":992,"description":993,"org":994,"tags":995,"stars":20,"repoUrl":21,"updatedAt":1004},"icd10-cm-skill","extract ICD-10-CM diagnosis codes","Extract billable ICD-10-CM diagnosis codes from a clinical note the way a professional coder builds the claim. Use when users say \"code this encounter\", \"assign ICD-10 codes\", \"what diagnosis codes apply\", \"code this chart\", or when turning clinical documentation into claim-ready diagnosis codes.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[996,999,1000,1001],{"name":997,"slug":998,"type":16},"Coding","coding",{"name":14,"slug":15,"type":16},{"name":987,"slug":988,"type":16},{"name":1002,"slug":1003,"type":16},"Medical Necessity","medical-necessity","2026-06-19T09:34:31.578353",9,{"items":1007,"total":1186},[1008,1029,1043,1053,1072,1085,1102,1122,1136,1149,1157,1170],{"slug":1009,"name":1009,"fn":1010,"description":1011,"org":1012,"tags":1013,"stars":1026,"repoUrl":1027,"updatedAt":1028},"algorithmic-art","create algorithmic art with p5.js","Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1014,1017,1020,1023],{"name":1015,"slug":1016,"type":16},"Creative","creative",{"name":1018,"slug":1019,"type":16},"Design","design",{"name":1021,"slug":1022,"type":16},"Generative Art","generative-art",{"name":1024,"slug":1025,"type":16},"JavaScript","javascript",161831,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fskills","2026-04-06T17:56:15.455818",{"slug":1030,"name":1030,"fn":1031,"description":1032,"org":1033,"tags":1034,"stars":1026,"repoUrl":1027,"updatedAt":1042},"brand-guidelines","apply Anthropic brand colors and typography","Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1035,1038,1039],{"name":1036,"slug":1037,"type":16},"Branding","branding",{"name":1018,"slug":1019,"type":16},{"name":1040,"slug":1041,"type":16},"Typography","typography","2026-04-06T17:56:05.042852",{"slug":1044,"name":1044,"fn":1045,"description":1046,"org":1047,"tags":1048,"stars":1026,"repoUrl":1027,"updatedAt":1052},"canvas-design","create posters and visual art as PNG or PDF","Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1049,1050,1051],{"name":1015,"slug":1016,"type":16},{"name":1018,"slug":1019,"type":16},{"name":941,"slug":942,"type":16},"2026-04-06T17:56:03.794732",{"slug":1054,"name":1054,"fn":1055,"description":1056,"org":1057,"tags":1058,"stars":1026,"repoUrl":1027,"updatedAt":1071},"claude-api","build apps with the Claude API","Reference for the Claude API \u002F Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration.\nTRIGGER — read BEFORE opening the target file; don't skip because it \"looks like a one-liner\" — whenever: the prompt names Claude\u002FAnthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing\u002Fmodel choice\u002Flimits\u002Fcaching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent\u002FMCP\u002Ftool-definition\u002Fmulti-agent\u002FRAG\u002FLLM-judge\u002Fcomputer-use; generate\u002Fsummarize\u002Fextract\u002Fclassify\u002Frewrite\u002Fconverse over NL; debugging refusals\u002Fcutoffs\u002Fstreaming\u002Ftool-calls\u002Ftokens).\nSKIP only when another provider is being worked on (overrides all triggers): OpenAI\u002FGPT\u002FGemini\u002FLlama\u002FMistral\u002FCohere\u002FOllama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST if no provider named — don't Read the file).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1059,1062,1063,1066,1068],{"name":1060,"slug":1061,"type":16},"Agents","agents",{"name":9,"slug":8,"type":16},{"name":1064,"slug":1065,"type":16},"Anthropic SDK","anthropic-sdk",{"name":1067,"slug":1054,"type":16},"Claude API",{"name":1069,"slug":1070,"type":16},"LLM","llm","2026-07-28T05:36:08.213335",{"slug":1073,"name":1073,"fn":1074,"description":1075,"org":1076,"tags":1077,"stars":1026,"repoUrl":1027,"updatedAt":1084},"doc-coauthoring","co-author documentation and technical specs","Guide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1078,1081],{"name":1079,"slug":1080,"type":16},"Documentation","documentation",{"name":1082,"slug":1083,"type":16},"Technical Writing","technical-writing","2026-04-06T17:56:14.18897",{"slug":939,"name":939,"fn":1086,"description":1087,"org":1088,"tags":1089,"stars":1026,"repoUrl":1027,"updatedAt":1101},"create and edit Word documents","Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files) or Word templates (.dotx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', '.dotx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx or .dotx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1090,1091,1092,1095,1098],{"name":918,"slug":919,"type":16},{"name":938,"slug":939,"type":16},{"name":1093,"slug":1094,"type":16},"Office","office",{"name":1096,"slug":1097,"type":16},"Templates","templates",{"name":1099,"slug":1100,"type":16},"Word","word","2026-07-18T05:16:23.136271",{"slug":1103,"name":1103,"fn":1104,"description":1105,"org":1106,"tags":1107,"stars":1026,"repoUrl":1027,"updatedAt":1121},"frontend-design","design production-grade frontend interfaces","Guidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1108,1109,1112,1115,1118],{"name":1018,"slug":1019,"type":16},{"name":1110,"slug":1111,"type":16},"Frontend","frontend",{"name":1113,"slug":1114,"type":16},"React","react",{"name":1116,"slug":1117,"type":16},"Tailwind CSS","tailwind-css",{"name":1119,"slug":1120,"type":16},"UI Components","ui-components","2026-04-06T17:56:16.723469",{"slug":1123,"name":1123,"fn":1124,"description":1125,"org":1126,"tags":1127,"stars":1026,"repoUrl":1027,"updatedAt":1135},"internal-comms","write internal company communications","A set of resources to help me write all kinds of internal communications, using the formats that my company likes to use. Claude should use this skill whenever asked to write some sort of internal communications (status reports, leadership updates, 3P updates, company newsletters, FAQs, incident reports, project updates, etc.).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1128,1131,1132],{"name":1129,"slug":1130,"type":16},"Communications","communications",{"name":1096,"slug":1097,"type":16},{"name":1133,"slug":1134,"type":16},"Writing","writing","2026-04-06T17:56:20.695522",{"slug":1137,"name":1137,"fn":1138,"description":1139,"org":1140,"tags":1141,"stars":1026,"repoUrl":1027,"updatedAt":1148},"mcp-builder","build MCP servers","Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node\u002FTypeScript (MCP SDK).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1142,1143,1144,1145],{"name":1060,"slug":1061,"type":16},{"name":954,"slug":955,"type":16},{"name":1069,"slug":1070,"type":16},{"name":1146,"slug":1147,"type":16},"MCP","mcp","2026-04-06T17:56:10.357665",{"slug":942,"name":942,"fn":1150,"description":1151,"org":1152,"tags":1153,"stars":1026,"repoUrl":1027,"updatedAt":1156},"read edit and manipulate PDF files","Use this skill whenever the user wants to do anything with PDF files. This includes reading or extracting text\u002Ftables from PDFs, combining or merging multiple PDFs into one, splitting PDFs apart, rotating pages, adding watermarks, creating new PDFs, filling PDF forms, encrypting\u002Fdecrypting PDFs, extracting images, and OCR on scanned PDFs to make them searchable. If the user mentions a .pdf file or asks to produce one, use this skill.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1154,1155],{"name":918,"slug":919,"type":16},{"name":941,"slug":942,"type":16},"2026-04-06T17:56:02.483316",{"slug":1158,"name":1158,"fn":1159,"description":1160,"org":1161,"tags":1162,"stars":1026,"repoUrl":1027,"updatedAt":1169},"pptx","create and edit PowerPoint presentations","Use this skill any time a .pptx or .potx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx or .potx file (even if the extracted content will be used elsewhere, like in an email or summary); editing, modifying, or updating existing presentations; combining or splitting slide files; working with templates (.potx), layouts, speaker notes, or comments. Trigger whenever the user mentions \"deck,\" \"slides,\" \"presentation,\" or references a .pptx or .potx filename, regardless of what they plan to do with the content afterward. If a .pptx or .potx file needs to be opened, created, or touched, use this skill.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1163,1166],{"name":1164,"slug":1165,"type":16},"PowerPoint","powerpoint",{"name":1167,"slug":1168,"type":16},"Presentations","presentations","2026-07-18T05:16:24.1471",{"slug":1171,"name":1171,"fn":1172,"description":1173,"org":1174,"tags":1175,"stars":1026,"repoUrl":1027,"updatedAt":1185},"skill-creator","create and optimize agent skills","Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1176,1177,1178,1181,1184],{"name":1060,"slug":1061,"type":16},{"name":1079,"slug":1080,"type":16},{"name":1179,"slug":1180,"type":16},"Evals","evals",{"name":1182,"slug":1183,"type":16},"Performance","performance",{"name":1082,"slug":1083,"type":16},"2026-04-19T06:45:40.804",490]