Clinical Note Summarization

1. Load system prompt: prompt.md enforces no hallucinations + data gap surfacing.
2. Normalize input: Pre-clean vitals, labs, and timeline context when available.
3. Generate summary: Call preferred LLM (OpenAI, Anthropic, Gemini, OSS) using usage.py as a template.
4. Validate: Cross-check extracted values vs. source text and ensure contradictions/missing data are spelled out.
5. Deliver output: Provide markdown + JSON as required and log PHI handling steps.

Guardrails

• Never invent findings; state "not provided" explicitly.
• Mark outputs as documentation support only—not clinical decisions.
• Strip/re-mask PHI before storing prompts/responses.

References

• For detailed schema, guardrails, and integration snippets see README.md, prompt.md, and usage.py.