symptomguide ai

The problem

Users describing health concerns to AI often receive either dangerous medical advice or vague, unhelpful disclaimers. Standard LLMs lack the structured triage protocols needed to safely distinguish between mild discomfort and life-threatening emergencies.

What it does

• Identifies life-threatening "red flags" immediately using a structured triage protocol before performing any analysis.
• Conducts a clinical-grade anamnesis using the SOCRATES method to gather site, onset, and severity data.
• Provides 2-4 plausible hypotheses with supporting and contradicting evidence instead of a single diagnostic claim.
• Categorizes every case into one of five urgency levels, from immediate emergency to self-monitoring.
• Generates a structured medical report for the user to share with their actual healthcare provider.

Why this beats prompting it yourself

This skill implements strict safety boundaries that prevent the AI from prescribing medication or making definitive diagnoses. It uses a multi-step clinical reasoning chain that forces emergency checks first, preventing the "hallucination" of safety in critical situations.

Use cases

• Initial orientation for users unsure if a symptom requires an ER visit or a GP appointment.
• Structured preparation for a doctor's consultation to ensure no clinical detail is forgotten.
• Safe health guidance for vulnerable populations requiring threshold escalations, such as infants or the elderly.
• Health app integration for preliminary symptom gathering before teleconsultation.

Known limitations

Cannot interpret lab results or medical imaging. Does not handle mental health crises or veterinary cases. Minimum age for automated orientation is 3 months.