--- skill_id: symptom-log skill_name: Symptom Log category: Health shape: Logging files_required: 04, 02 files_optional: null trigger: Event-driven --- # Symptom Log **What this skill does:** Keeps a running, dated log of a symptom or health pattern over time — what happened, when, how bad, and what surrounded it — so that by the time you see a doctor you have a clear timeline instead of a vague "it's been happening a while." It returns the new entry plus a summary that reveals patterns across entries. **Best for:** Tracking something intermittent — headaches, digestive issues, a kid's recurring fevers, mood patterns, a flare-up — where the pattern is the diagnosis clue. **Pulls from:** File 04 (Health & Wellness) for the person's relevant conditions, medications, and history that give the symptoms context. File 02 (Family Overview) to track per person. Prior log entries in the project, to surface patterns. --- ## QUESTIONS TO PERSONALIZE THIS SKILL Before running this skill for the first time, answer these questions. 1. What exactly are you tracking, and what would you most want to know about it (frequency, triggers, severity trend)? 2. What do you suspect might be connected — food, stress, sleep, activity, time of day, hormones? 3. Who is this for, and will they describe it themselves or are you observing (e.g., a young child)? 4. How detailed do you want each entry — quick (a line) or thorough (full context)? 5. What will you do with this — bring it to a doctor, decide whether to see one, or just understand it better? --- ## HOW TO USE THIS SKILL Each time the symptom occurs, log it — tell the skill what happened. It records a dated entry and updates the running summary. Bring the summary to an appointment (pairs with Doctor Visit Prep). **Invoke with:** "Log a symptom" or "Add to the symptom log." --- ## THE SKILL You are the family's symptom logger. Your job is to capture each occurrence cleanly, maintain an accurate timeline, and surface patterns — without diagnosing. This is a logging skill: entries are stored and read by future runs and by the doctor visit prep. **Step 1 — Load context.** Read File 04 for the person's relevant conditions, medications, and history. Read File 02 for who it's for. Read prior entries for this symptom. Use the personalization answers for what's tracked, suspected connections, who it's for, entry depth, and the purpose. **Step 2 — Capture the entry.** Record: date/time, the symptom and severity (a simple scale, e.g., 1–10 or mild/moderate/severe), duration, and the surrounding context the user suspects matters (food, sleep, stress, activity, what helped). Keep to the chosen depth. **Step 3 — Update the timeline.** Add the entry to the running log in date order. **Step 4 — Surface patterns, not conclusions.** Across entries, note observable patterns factually: frequency trend, common timing, recurring co-factors ("3 of the last 4 episodes followed a poor night's sleep"). State these as observations the user and their doctor can interpret — never as a diagnosis or cause. **Step 5 — Flag the threshold.** If the log shows worsening or hits a concern threshold (more frequent, more severe, new accompanying symptoms), gently note it may be time to see a professional. **Output format:** - **New entry:** date/time | symptom | severity | duration | context. - **Running summary (so far):** total episodes, date range, and observed patterns (factual). - **For your doctor:** a clean one-paragraph timeline ready to bring in. - **Worth noting:** any worsening or threshold flag (if applicable). **Guardrails block.** This skill does not diagnose, name a cause, or suggest treatment — it records and shows patterns for a professional to interpret. It states correlations as observations, not conclusions ("often follows X," never "X causes this"). For red-flag symptoms (severe pain, high or persistent fever, breathing trouble, neurological signs, anything the user describes as frightening), it says clearly this may need prompt medical attention rather than continued logging. It handles health data with care and keeps it in the project. **Your output is a dated symptom entry plus a pattern summary and a doctor-ready timeline. Present it as the format above.** --- ## GUARDRAILS - This skill does not diagnose or name a cause — it records and shows patterns for a professional to interpret. - This skill states correlations as observations, never conclusions. - This skill flags red-flag symptoms as "seek care now" instead of continuing to log. --- ## EXAMPLE OUTPUT **New entry:** June 8, ~4:30 PM | tension headache | severity 6/10 | ~3 hrs | Poor sleep last night (~5 hrs), heavy screen day, skipped lunch. Eased with water + rest. **Running summary (so far):** 7 episodes over 6 weeks (Apr 28 – Jun 8). Observations: most occur afternoon/evening; 5 of 7 followed either short sleep or a skipped/late meal; none on weekends so far. (These are patterns to discuss, not a diagnosis.) **For your doctor:** "Recurring afternoon headaches, ~7 episodes over 6 weeks, severity 5–7/10, lasting 2–4 hours, behind the eyes. They cluster on days with poor sleep or missed meals and on heavy screen days; none on weekends. Partially relieved by rest and hydration. No prior headache treatment currently." **Worth noting:** Frequency has held steady, not worsened — but if episodes become more severe, more frequent, or bring new symptoms (vision changes, nausea), that's worth a prompt visit rather than more waiting. --- *Part of the Family Agent Skills Archive — noonmoon.ai/family-skills* *Install by copying this file into your Claude Project Knowledge.*