Best for
- Use when: (1) Evaluating a patient's health record for coding issues (deprecated code systems, invalid codes, missing required fields), (2) Presenting issues in plain language with clinical impact, (3) Applying patient-…
aks129/HealthClawGuardrails/skills/curatr/SKILL.md
HealthClaw Curatr (healthclaw.io) — patient-facing FHIR data quality evaluation and correction. Use when: (1) Evaluating a patient's health record for coding issues (deprecated code systems, invalid codes, missing required fields), (2) Presenting issues in plain language with clinical impact, (3) Applying patient-approved corrections with full Provenance tracking, (4) Preparing a structured correction request for the patient's healthcare provider. Supports FHIR R4 US Core v9 resources: Condition
Decision brief
A patient-facing data quality skill that evaluates FHIR health records for coding issues, presents them in plain language, and lets the patient decide how to fix them.
Compatibility matrix
| Platform | Status | Evidence | What to check |
|---|---|---|---|
| Codex | Not declared | No explicit evidence | Portability before use |
| Claude Code | Not declared | No explicit evidence | Portability before use |
| Cursor | Not declared | No explicit evidence | Portability before use |
| Gemini CLI | Not declared | No explicit evidence | Portability before use |
Installation
The source command is displayed only when detected. A safe inspection prompt is always available so your agent can explain every action before execution.
npx skills add https://github.com/aks129/HealthClawGuardrails --skill "skills/curatr"Inspect the Agent Skill "curatr" from https://github.com/aks129/HealthClawGuardrails/blob/b94ae9ea95695d5755e22b02ef7b7385f0e3aee7/skills/curatr/SKILL.md at commit b94ae9ea95695d5755e22b02ef7b7385f0e3aee7. List every install step, command, network request, credential, file read/write, external action, and rollback step. Explain whether it fits my task. Do not install or execute anything until I approve.
Workflow
Apply patient-approved fixes to a FHIR resource. Creates a linked Provenance record.
This skill runs on top of the HealthClaw Guardrails stack. See the fhir-r6-guardrails skill for installation instructions.
For all other resource types, Curatr runs a generic coding scan checking all coding[] elements for deprecated systems.
Deprecated systems: ICD-9-CM codes (retired October 2015) flagged as critical
Deprecated systems: ICD-9-CM codes (retired October 2015) flagged as critical
Permission review
The documentation includes network, browsing, or remote request actions.
"new_value": "http://hl7.org/fhir/sid/icd-10-cm"The documentation asks the agent to run terminal commands or scripts.
docker-compose up -d --buildThe documentation includes network, browsing, or remote request actions.
# POST http://localhost:3001/mcp (Streamable HTTP, preferred)Evidence record
| Signal | Value | Evidence type | Meaning |
|---|---|---|---|
| Quality score | 84/100 | Computed | Documentation, specificity, maintenance, and trust rules |
| Repository stars | 27 | Source | Repository attention, not individual Skill quality |
| Compatibility | 0 platforms | Source | Declared in the catalog source record |
| Usage guide | automated source guide | Editorial | Generated or reviewed according to the visible evidence level |
Pinned source
A patient-facing data quality skill that evaluates FHIR health records for coding issues, presents them in plain language, and lets the patient decide how to fix them.
Your health data belongs to you. Curatr helps you understand what's in it, spot problems, and take action — whether that's correcting your personal record or requesting your provider fix the source.
| Resource | Checks |
|---|---|
| Condition | ICD-9 detection, ICD-10-CM/SNOMED validity, clinicalStatus, verificationStatus |
| AllergyIntolerance | clinicalStatus, patient link, allergen code (RxNorm/SNOMED) |
| MedicationRequest | status, intent, medication code (RxNorm) |
| Immunization | status, vaccineCode (CVX/SNOMED), occurrenceDateTime |
| Procedure | status, procedure code (SNOMED/CPT) |
| DiagnosticReport | status, report code (LOINC) |
For all other resource types, Curatr runs a generic coding scan checking all coding[]
elements for deprecated systems.
| Service | Systems Validated |
|---|---|
| tx.fhir.org — HL7 public FHIR terminology server | SNOMED CT, LOINC |
| NLM Clinical Tables API | ICD-10-CM |
| RXNAV API | RxNorm |
curatr.evaluate (read-only)Evaluate a FHIR resource for data quality issues.
Input:
{
"resource_type": "Condition",
"resource_id": "cond-001"
}
Output: Issues list with plain-language descriptions, impact, and suggestions.
Each issue includes severity (critical / warning / info / suggestion), field_path,
plain_language, impact, and suggestion.
No step-up required. Safe to call at any time.
curatr.apply_fix (write — requires step-up + human confirmation)Apply patient-approved fixes to a FHIR resource. Creates a linked Provenance record.
Input:
{
"resource_type": "Condition",
"resource_id": "cond-001",
"fixes": [
{
"field_path": "Condition.code.coding[0].system",
"new_value": "http://hl7.org/fhir/sid/icd-10-cm"
},
{
"field_path": "Condition.code.coding[0].code",
"new_value": "E11.9"
},
{
"field_path": "Condition.code.coding[0].display",
"new_value": "Type 2 diabetes mellitus without complications"
}
],
"patient_intent": "Updating from retired ICD-9 to ICD-10-CM equivalent"
}
Requires: X-Step-Up-Token header + X-Human-Confirmed: true
Output: Updated resource + linked Provenance resource documenting the change.
Always follow this sequence — never apply fixes without patient approval:
1. Call curatr.evaluate to get issues
2. For each issue, present:
- issue.title (bold headline)
- issue.plain_language (what the problem is, in plain English)
- issue.impact (why it matters to the patient)
- issue.suggestion (what to do about it)
3. Ask the patient which fixes they approve
4. Confirm their intent in their own words (used in Provenance)
5. Call curatr.apply_fix with ONLY the approved fixes
6. Confirm completion and show the Provenance record summary
Curatr found 2 issues in your Diabetes condition record:
─────────────────────────────────────────────────────────
CRITICAL: Outdated code system
Your record uses ICD-9-CM — a code system retired in October 2015.
Most US health systems and insurers no longer accept ICD-9 codes.
Impact: This record may not match your condition when shared with
providers or insurance systems, potentially causing delays.
Suggested fix: Update to ICD-10-CM code E11.9 — "Type 2 diabetes
mellitus without complications"
─────────────────────────────────────────────────────────
INFO: Missing verification status
No indication of whether this diagnosis has been confirmed.
Impact: Without a verification status, AI tools and care summaries
may not know how to weight this condition.
Suggested fix: Add verificationStatus = "confirmed" (or "provisional"
if still being evaluated by your provider)
─────────────────────────────────────────────────────────
What would you like to do?
[A] Fix both in my personal health record
[B] Fix the code system only
[C] Prepare a correction request for my provider
[D] Keep as-is for now
| Level | Meaning | Example |
|---|---|---|
critical | Code system deprecated or structurally broken | ICD-9 code in 2026 |
warning | Code invalid or display name wrong | Invalid clinicalStatus value |
info | Recommended field missing | No verificationStatus |
suggestion | A better value exists | Display name differs from canonical |
Every fix creates a FHIR Provenance resource linked to the updated resource:
UPDATE / revise from HL7 v3 DataOperationPATADMIN (patient administration)The original source record is preserved in the AuditEvent trail. Provenance does not erase history — it adds attribution.
If the patient wants their provider to fix the source record:
This skill runs on top of the HealthClaw Guardrails stack.
See the fhir-r6-guardrails skill for installation instructions.
# Curatr requires the same stack — no additional services needed
export STEP_UP_SECRET=$(openssl rand -hex 32)
docker-compose up -d --build
# MCP tools available at:
# POST http://localhost:3001/mcp (Streamable HTTP, preferred)
# GET http://localhost:3001/sse (SSE, legacy)
Alternatives
coreyhaines31/marketingskills
When the user wants to plan, design, or implement an A/B test or experiment, or build a growth experimentation program. Also use when the user mentions "A/B test," "split test," "experiment," "test this change," "variant copy," "multivariate test," "hypothesis," "should I test this," "which version is better," "test two versions," "statistical significance," "how long should I run this test," "growth experiments," "experiment velocity," "experiment backlog," "ICE score," "experimentation program
alirezarezvani/claude-skills
App Store Optimization (ASO) toolkit for researching keywords, analyzing competitor rankings, generating metadata suggestions, and improving app visibility on Apple App Store and Google Play Store. Use when the user asks about ASO, app store rankings, app metadata, app titles and descriptions, app store listings, app visibility, or mobile app marketing on iOS or Android. Supports keyword research and scoring, competitor keyword analysis, metadata optimization, A/B test planning, launch checklist
dotnet/skills
Migrates .NET test projects from VSTest to Microsoft.Testing.Platform (MTP). Use when user asks to "migrate to MTP", "switch from VSTest", "enable Microsoft.Testing.Platform", "use MTP runner", set OutputType=Exe only for test projects in Directory.Build.props, or mentions EnableMSTestRunner, EnableNUnitRunner, or UseMicrosoftTestingPlatformRunner. USE FOR: MTP behavioral differences vs VSTest (exit code 8, zero tests discovered, --ignore-exit-code, TESTINGPLATFORM_EXITCODE_IGNORE); centralizing
JasonColapietro/suede-creator-skills
Suede-owned experimentation discipline for hypotheses, sample sizing, test duration, significance, and repeatable experiment programs. Use when comparing variants, deciding whether a result is reliable, or building an experiment backlog and cadence. NOT FOR: analytics instrumentation (use suede-analytics), post-click conversion diagnosis (use suede-site-alchemy), or writing the variant copy itself (use suede-copy).