Diagnose a low-energy health issue with an agent
I had a situation where I felt incredibly low energy and I knew something was wrong. Task: The user built a custom agent to analyze persistent low energy symptoms after four doctors dismissed concerns. The agent processed symptom inputs to recommend specific blood tests and specialist referrals. The human acted as the primary investigator, using the agent's output to guide medical consultations and validate findings against professional advice.; tools: custom-built or unnamed agent. Human role: The original post does not say how a person stayed involved.; limitations: The source is a secondary anecdotal post with no technical documentation of the agent's architecture or data sources. Medical advice from AI agents carries significant safety risks and should not replace professional diagnosis. The workflow relies on the user's ability to interpret AI suggestions and seek appropriate medical care, with no automated safeguards mentioned..
Tools
custom-built or unnamed agent
Published on Agent Work Atlas:
Overview
I had a situation where I felt incredibly low energy and I knew something was wrong.
Workflow
The user built a custom agent to analyze persistent low energy symptoms after four doctors dismissed concerns. The agent processed symptom inputs to recommend specific blood tests and specialist referrals. The human acted as the primary investigator, using the agent's output to guide medical consultations and validate findings against professional advice.
Reported result
The agent identified a solution that matched the diagnosis provided by a fifth doctor, who agreed with the AI's assessment after previous doctors dismissed the symptoms. No quantitative metrics or formal validation data are provided in the source. The outcome is described as the agent finding the correct medical path, but the specific diagnosis and treatment details are omitted.
Human role
The original post does not say how a person stayed involved.
Limitations
The source is a secondary anecdotal post with no technical documentation of the agent's architecture or data sources. Medical advice from AI agents carries significant safety risks and should not replace professional diagnosis. The workflow relies on the user's ability to interpret AI suggestions and seek appropriate medical care, with no automated safeguards mentioned.
Try it
Act as a health research assistant. I will provide a list of persistent symptoms and previous medical feedback. [replaceable inputs: symptom list, doctor feedback, personal health history]. Step 1: Analyze the symptoms to suggest potential underlying causes and specific blood tests or specialist referrals. Step 2: Generate a structured report with recommended next steps for medical consultation. Step 3: Do not provide a diagnosis or treatment plan. Output format: A markdown table with columns for 'Symptom', 'Potential Cause', 'Recommended Test/Specialist', and 'Rationale'. Measure success by the relevance of suggested tests to standard medical guidelines for the described symptoms. Handle missing inputs by asking for clarification on symptom duration and severity. Require my explicit confirmation before finalizing the report. Note: This is for informational purposes only and does not constitute medical advice.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.