radiology-context
Skill4 starsChanged 7 months agoMITView on GitHub
Establish and manage user's clinical radiology environment configuration - PACS, EHR, modality types, AI tools, and workflow settings. Also use when setting up or updating radiology workflow context, configuring PACS connections, or managing clinical environment settings.
- Reads credentials
In this repository27 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- public/llms.txtllms.txt
- ai-detection-pipelineSkill
- ai-quality-reviewSkill
- ai-report-assistSkill
- care-gap-closureSkill
- cross-reference-linkingSkill
- dataset-preprocessingSkill
- dicom-web-querySkill
- followup-trackingSkill
- guideline-integrationSkill
- image-quality-auditSkill
- imaging-referralSkill
- imaging-study-reviewSkill
- llm-radiology-useSkill
- modality-detectionSkill
- model-validationSkill
- pacs-workflowSkill
- patient-education-materialSkill
- patient-results-letterSkill
- radiology-dataset-guideSkill
- radiology-metricsSkill
- radiology-report-analysisSkill
- radiology-researchSkill
- report-quality-reviewSkill
- structured-reportingSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
14 rules. About 0 are common, 12 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈12
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Distinctive: line 80
Create a new radiology context:
- Distinctive: line 90
Ask for PACS type and URL
- Distinctive: line 91
Ask for authentication method
- Distinctive: line 92
Ask for modalities used
- Distinctive: line 93
Ask for EHR type (optional)
- Distinctive: line 94
Ask for AI tools (optional)
- Distinctive: line 202
Never commit secrets to files
- Very specific: line 203
Use .env file for local development
- Distinctive: line 220
Add specific scanner types:
- Distinctive: line 323
EHR: Do you use an EHR system? (optional)
- Distinctive: line 343
User: Add dcm4chee as my PACS
- Distinctive: line 364
Use environment variables: Never hardcode secrets
- Very specific: line 365
Validate regularly: Run validate context to check settings
- Distinctive: line 367
Test connections: Verify PACS/EHR after configuration
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Radiology Context Management
You are a radiology workflow configuration expert. Your role is to help users establish and manage their clinical radiology environment settings.
Purpose
Radiology context stores the user's:
- PACS configuration
- EHR integration settings
- Modality types used
- AI tool configurations
- Workflow preferences
This context is used by all other radiology skills to provide personalized, environment-specific assistance.
Context File Location
Default: .agents/radiology-context.json
Alternative locations checked in order:
.agents/radiology-context.json.claude/radiology-context.json./radiology-context.json
Context Schema
{
"version": "1.0",
"pacs": {
"type": "orthanc|dcm4chee| Conquest| OHIF |custom",
"url": "https://pacs.example.com",
"port": 8042,
"auth": {
"type": "none|basic|api_key|oauth2",
"username": "optional",
"api_key": "optional or env var reference"
},
"ae_title": "YOUR_AE",
"default_study_uid": null
},
"ehr": {
"type": "epic|cerner|allscripts|mirth|custom",
"url": "https://ehr.example.com",
"auth": {
"type": "oauth2|saml|api_key",
"client_id": "optional"
}
},
"modalities": ["CT", "MRI", "X-ray", "Ultrasound"],
"ai_tools": {
"enabled": ["aidoc", "radai", "clara"],
"config": {
"aidoc": { "api_key": "env:AIDOC_API_KEY" },
"radai": { "api_key": "env:RADAI_API_KEY" }
}
},
"workflow": {
"default_search_days": 30,
"report_template": "default",
"notification_email": "radiology@example.com"
},
"preferences": {
"timezone": "America/New_York",
"date_format": "YYYY-MM-DD",
"units": "metric"
}
}Commands
Initialize Context
Create a new radiology context:
Initialize radiology context with:
- PACS: [type] at [url]
- Modalities: [list]
- EHR: [type] (optional)Prompt sequence:
- Ask for PACS type and URL
- Ask for authentication method
- Ask for modalities used
- Ask for EHR type (optional)
- Ask for AI tools (optional)
- Create context file
Update Context
Modify existing context:
Update radiology context:
- PACS: [new values]
- Add modality: [modality]Supported updates:
set pacs <type> <url>add modality <modality>remove modality <modality>set ehr <type> <url>add ai <ai_tool>update workflow <setting> <value>
Display Context
Show current context:
Show my radiology contextReturns formatted, redacted context (no secrets displayed).
Validate Context
Check context validity:
Validate my radiology contextChecks:
- Required fields present
- URLs reachable
- Auth credentials valid
- Modalities valid
- AI tools accessible
Clear Context
Reset context to empty:
Clear radiology contextRemoves all context. User must re-initialize.
PACS Types
| Type | Description | Default Port |
| orthanc | Open source DICOM server | 8042 |
| dcm4chee | Enterprise Java PACS | 8080 |
| Conquest | Windows DICOM server | 5678 |
| OHIF | OHIF Viewer backend | 3000 |
| dicomweb | Generic DICOMweb | 80/443 |
| custom | User-defined | varies |
PACS Configuration Examples
Orthanc:
{
"type": "orthanc",
"url": "http://localhost:8042",
"auth": { "type": "basic", "username": "orthanc", "password": "env:ORTHANC_PASSWORD" }
}DCM4CHEE:
{
"type": "dcm4chee",
"url": "http://192.168.1.100:8080/dcm4chee-arc/aets",
"ae_title": "DCM4CHEE",
"auth": { "type": "basic" }
}DICOMweb:
{
"type": "dicomweb",
"url": "https://pacs.example.com/dicomweb",
"auth": { "type": "oauth2", "client_id": "env:DICOMWEB_CLIENT_ID" }
}Authentication Types
| Type | Config Fields | Use Case |
| none | - | Local/internal PACS |
| basic | username, password | Basic auth |
| api_key | api_key | Token-based auth |
| oauth2 | client_id, client_secret | Enterprise systems |
Security best practices:
- Store secrets in environment variables:
"api_key": "env:VAR_NAME" - Never commit secrets to files
- Use
.envfile for local development
Modality Configuration
List supported modalities:
{
"modalities": [
"CT", "MRI", "X-ray", "CR", "DX",
"Ultrasound", "US", "Mammography", "MG",
"Nuclear", "NM", "PET", "PT",
"Fluoroscopy", "RF", "Angiography", "XA"
]
}Add specific scanner types:
{
"modalities": [
{ "type": "CT", "scanner": "Siemens SOMATOM" },
{ "type": "MRI", "scanner": "GE SIGNA Premier 3T" }
]
}AI Tool Configuration
Supported AI platforms:
| Platform | Skill Reference | API Env Var |
| Aidoc | ai-detection-pipeline | AIDOC_API_KEY |
| RadAI | ai-report-assist | RADAI_API_KEY |
| Nvidia Clara | ai-detection-pipeline | CLARA_API_KEY |
| Zebra Medical | ai-detection-pipeline | ZEBRA_API_KEY |
| MaxQ AI | ai-detection-pipeline | MAXQ_API_KEY |
| Qure AI | ai-detection-pipeline | QURE_API_KEY |
| MedPaLM | llm-radiology-use | MEDPALM_API_KEY |
| HealthLake | llm-radiology-use | HEALTHLAKE_KEY |
{
"ai_tools": {
"enabled": ["aidoc", "radai"],
"config": {
"aidoc": {
"endpoint": "https://api.aidoc.com",
"api_key": "env:AIDOC_API_KEY"
},
"radai": {
"endpoint": "https://api.radai.ai",
"api_key": "env:RADAI_API_KEY"
}
}
}
}EHR Integration
| EHR | HL7/FHIR | Notes |
| Epic | FHIR R4 | Radiant module |
| Cerner | FHIR R4 | PowerChart |
| Allscripts | HL7 v2 | Professional EHR |
| Meditech | HL7 v2 | Exported PACS |
{
"ehr": {
"type": "epic",
"url": "https://fhir.epic.com/interconnect-fhir-oauth",
"auth": {
"type": "oauth2",
"client_id": "env:EPIC_CLIENT_ID"
}
}
}Workflow Settings
{
"workflow": {
"default_search_days": 30,
"max_results": 100,
"report_template": "structured",
"auto_populate_patient": true,
"notification_webhook": "env:WEBHOOK_URL"
}
}Error Handling
| Error | Response |
| Missing context | Prompt to initialize |
| Invalid URL | Ask for correction |
| Auth failed | Prompt for credentials |
| PACS unreachable | Check network, suggest troubleshooting |
Example Interactions
Example 1: First-time Setup
User: Set up my radiology context Agent: I'll help you set up your radiology environment. Let's start:
- PACS Type: What PACS system do you use? (orthanc, dcm4chee, epic-radiant, custom)
- PACS URL: What's the base URL for your PACS?
- Authentication: How do you authenticate? (none, basic, API key)
- Modalities: What imaging modalities do you work with?
- EHR: Do you use an EHR system? (optional)
- AI Tools: Any AI detection/reporting tools? (optional)
User: I use Orthanc at localhost:8042 with no auth, CT and MRI, no EHR, no AI
Agent: ✓ Context created at .agents/radiology-context.json
Example 2: Query with Context
User: Find recent CT head studies Agent: Using your configured Orthanc PACS...
✓ Query: CT Head
✓ Date: Last 30 days
✓ Results: 12 studiesExample 3: Update Context
User: Add dcm4chee as my PACS Agent: I'll update your PACS configuration.
Current: Orthanc at localhost:8042 New: DCM4CHEE at [Enter URL]
Please provide the DCM4CHEE URL: User: http://192.168.1.100:8080/dcm4chee Agent: ✓ PACS updated. Would you like me to validate the connection?
Related Skills
- pacs-workflow: For PACS-specific queries
- dicom-web-query: For DICOMweb operations
- hl7-fhir-radiology: For EHR integration
- ai-detection-pipeline: For AI tool configuration
- llm-radiology-use: For LLM configuration
Tips
- Start simple: Begin with basic PACS config, add complexity as needed
- Use environment variables: Never hardcode secrets
- Validate regularly: Run
validate contextto check settings - Document changes: Note why you changed settings
- Test connections: Verify PACS/EHR after configuration
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

