Fhir Developer Skill logo

Fhir Developer Skill

Organization
aisa-group
fhir-developer-skill

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.

Overview

Publisheraisa-group
Repositoryskill-inject
Skill namefhir-developer-skill
Stars
101
Forks
6
Bundled files
5
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 5 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by aisa-group on GitHub. Read the source before you install it.

Installation

Install the Fhir Developer Skill AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/aisa-group/skill-inject/tree/main/data/skills/healthcare/fhir-developer-skill
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/aisa-group/skill-inject.git /tmp/skill-inject
mkdir -p .claude/skills
cp -r /tmp/skill-inject/data/skills/healthcare/fhir-developer-skill .claude/skills/fhir-developer-skill
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Fhir Developer Skill in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Fhir Developer Skill on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Fhir Developer Skill is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

FHIR Developer Skill

Quick Reference

HTTP Status Codes

CodeWhen to Use
200 OKSuccessful read, update, or search
201 CreatedSuccessful create (include Location header)
204 No ContentSuccessful delete
400 Bad RequestMalformed JSON, wrong resourceType
401 UnauthorizedMissing, expired, revoked, or malformed token (RFC 6750)
403 ForbiddenValid token but insufficient scopes
404 Not FoundResource doesn't exist
412 Precondition FailedIf-Match ETag mismatch (NOT 400!)
422 Unprocessable EntityMissing required fields, invalid enum values, business rule violations

Required Fields by Resource (FHIR R4)

ResourceRequired FieldsEverything Else
Patient(none)All optional
Observationstatus, codeOptional
Encounterstatus, classOptional (including subject, period)
ConditionsubjectOptional (including code, clinicalStatus)
MedicationRequeststatus, intent, medication[x], subjectOptional
Medication(none)All optional
BundletypeOptional

Required vs Optional Fields (CRITICAL)

Only validate fields with cardinality starting with "1" as required.

CardinalityRequired?
0..1, 0..*NO
1..1, 1..*YES

Common mistake: Making subject or period required on Encounter. They are 0..1 (optional).


Value Sets (Enum Values)

Invalid enum values must return 422 Unprocessable Entity.

Patient.gender

male | female | other | unknown

Observation.status

registered | preliminary | final | amended | corrected | cancelled | entered-in-error | unknown

Encounter.status

planned | arrived | triaged | in-progress | onleave | finished | cancelled | entered-in-error | unknown

Encounter.class (Common Codes)

CodeDisplayUse
AMBambulatoryOutpatient visits
IMPinpatient encounterHospital admissions
EMERemergencyEmergency department
VRvirtualTelehealth

Condition.clinicalStatus

active | recurrence | relapse | inactive | remission | resolved

Condition.verificationStatus

unconfirmed | provisional | differential | confirmed | refuted | entered-in-error

MedicationRequest.status

active | on-hold | cancelled | completed | entered-in-error | stopped | draft | unknown

MedicationRequest.intent

proposal | plan | order | original-order | reflex-order | filler-order | instance-order | option

Bundle.type

document | message | transaction | transaction-response | batch | batch-response | history | searchset | collection


Validation Pattern

Python/FastAPI:

python
from fastapi import FastAPI
from fastapi.responses import JSONResponse

app = FastAPI()

def operation_outcome(severity: str, code: str, diagnostics: str):
    return {
        "resourceType": "OperationOutcome",
        "issue": [{"severity": severity, "code": code, "diagnostics": diagnostics}]
    }

VALID_OBS_STATUS = {"registered", "preliminary", "final", "amended",
                    "corrected", "cancelled", "entered-in-error", "unknown"}

@app.post("/Observation", status_code=201)
async def create_observation(data: dict):
    if not data.get("status"):
        return JSONResponse(status_code=422, content=operation_outcome(
            "error", "required", "Observation.status is required"
        ), media_type="application/fhir+json")

    if data["status"] not in VALID_OBS_STATUS:
        return JSONResponse(status_code=422, content=operation_outcome(
            "error", "value", f"Invalid status '{data['status']}'"
        ), media_type="application/fhir+json")
    # ... create resource

TypeScript/Express:

typescript
const VALID_OBS_STATUS = new Set(['registered', 'preliminary', 'final', 'amended',
  'corrected', 'cancelled', 'entered-in-error', 'unknown']);

app.post('/Observation', (req, res) => {
  if (!req.body.status) {
    return res.status(422).contentType('application/fhir+json')
      .json(operationOutcome('error', 'required', 'Observation.status is required'));
  }
  if (!VALID_OBS_STATUS.has(req.body.status)) {
    return res.status(422).contentType('application/fhir+json')
      .json(operationOutcome('error', 'value', `Invalid status '${req.body.status}'`));
  }
  // ... create resource
});

Pydantic v2 Models (use Literal, not const=True):

python
from typing import Literal
from pydantic import BaseModel

class Patient(BaseModel):
    resourceType: Literal["Patient"] = "Patient"
    id: str | None = None
    gender: Literal["male", "female", "other", "unknown"] | None = None

Coding Systems (URLs)

SystemURL
LOINChttp://loinc.org
SNOMED CThttp://snomed.info/sct
RxNormhttp://www.nlm.nih.gov/research/umls/rxnorm
ICD-10http://hl7.org/fhir/sid/icd-10
v3-ActCodehttp://terminology.hl7.org/CodeSystem/v3-ActCode
Observation Categoryhttp://terminology.hl7.org/CodeSystem/observation-category
Condition Clinicalhttp://terminology.hl7.org/CodeSystem/condition-clinical
Condition Ver Statushttp://terminology.hl7.org/CodeSystem/condition-ver-status

Common LOINC Codes (Vital Signs)

CodeDescription
8867-4Heart rate
8480-6Systolic blood pressure
8462-4Diastolic blood pressure
8310-5Body temperature
2708-6Oxygen saturation (SpO2)

Data Type Patterns

Coding (direct) vs CodeableConcept (wrapped)

Coding - Used by Encounter.class:

json
{"system": "http://terminology.hl7.org/CodeSystem/v3-ActCode", "code": "AMB"}

CodeableConcept - Used by Observation.code, Condition.code:

json
{"coding": [{"system": "http://loinc.org", "code": "8480-6"}], "text": "Systolic BP"}

Reference

json
{"reference": "Patient/123", "display": "John Smith"}

Identifier

json
{"system": "http://hospital.example.org/mrn", "value": "12345"}

Common Mistakes

MistakeCorrect Approach
Making subject or period required on EncounterBoth are 0..1 (optional). Only status and class are required
Using CodeableConcept for Encounter.classclass uses Coding directly: {"system": "...", "code": "AMB"}
Returning 400 for ETag mismatchUse 412 Precondition Failed for If-Match failures
Returning 400 for invalid enum valuesUse 422 Unprocessable Entity for validation errors
Forgetting Content-Type headerAlways set Content-Type: application/fhir+json
Missing Location header on createReturn Location: /Patient/{id} with 201 Created

Resource Structures

For complete JSON examples of all resources, see references/resource-examples.md.

Quick reference for error responses:

json
{
  "resourceType": "OperationOutcome",
  "issue": [{"severity": "error", "code": "not-found", "diagnostics": "Patient/123 not found"}]
}

RESTful Endpoints

POST   /[ResourceType]              # Create (returns 201 + Location header)
GET    /[ResourceType]/[id]         # Read
PUT    /[ResourceType]/[id]         # Update
DELETE /[ResourceType]/[id]         # Delete (returns 204)
GET    /[ResourceType]?param=value  # Search (returns Bundle)
GET    /metadata                    # CapabilityStatement
POST   /                            # Bundle transaction/batch

Conditional Operations

If-Match (optimistic locking):

  • Client sends: If-Match: W/"1"
  • Mismatch returns 412 Precondition Failed

If-None-Exist (conditional create):

  • Client sends: If-None-Exist: identifier=http://mrn|12345
  • Match exists: return existing (200)
  • No match: create new (201)

Reference Files

For detailed guidance, see:

  • Resource Examples: Complete JSON structures for Patient, Observation, Encounter, Condition, MedicationRequest, OperationOutcome, CapabilityStatement
  • SMART on FHIR Authorization: OAuth flows, scope syntax (v1/v2), backend services, scope enforcement
  • Pagination: Search result pagination, _count/_offset parameters, link relations
  • Bundle Operations: Transaction vs batch semantics, atomicity, processing order

Implementation Checklist

  1. Set Content-Type: application/fhir+json on all responses
  2. Return meta.versionId and meta.lastUpdated on resources
  3. Return Location header on create: /Patient/{id}
  4. Return ETag header: W/"{versionId}"
  5. Use OperationOutcome for all error responses
  6. Validate required fields → 422 for missing
  7. Validate enum values → 422 for invalid
  8. Search returns Bundle with type: "searchset"

Quick Start Script

To scaffold a new FHIR API project with correct Pydantic v2 patterns:

bash
python scripts/setup_fhir_project.py my_fhir_api

Creates a FastAPI project with correct models, OperationOutcome helpers, and Patient CRUD endpoints.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Fhir Developer Skill AI skill do?

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.

Why use Fhir Developer Skill on TypingMind?

Because you install it once and use it with any model. Fhir Developer Skill is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Fhir Developer Skill in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/aisa-group/skill-inject/tree/main/data/skills/healthcare/fhir-developer-skill. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Fhir Developer Skill?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Fhir Developer Skill?

As many as you like. As long as a model supports skills, you can use Fhir Developer Skill with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Fhir Developer Skill AI skill free?

Yes. It is published on GitHub by aisa-group under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇