daily-log

name: daily-log description: Create or update a daily log. Triggers on "log today", "daily log", "log", "what happened today", "record today", or any request to create/update a daily entry.

Daily Log

Create or update a daily log from natural conversation.

Workflow

digraph log {
  "Determine date" -> "Check existing log";
  "Check existing log" -> "Fetch data (if integrations)";
  "Fetch data (if integrations)" -> "Conversational intake";
  "Conversational intake" -> "Write YAML + Notes";
  "Write YAML + Notes" -> "AI Observations";
  "AI Observations" -> "Save";
}

1. Determine Date

• Default: today
• If user specifies a date → use that
• If user says "yesterday" → use yesterday's date

2. Check Existing Log

Look for: logs/daily/YYYY-MM/YYYY-MM-DD.md

• If exists: Load it, offer to update/add to it
• If not: Create from logs/daily/TEMPLATE.md
• Create month folder if needed: logs/daily/YYYY-MM/

3. Fetch Data (if integrations configured)

Check CLAUDE.md for configured integrations (wearable commands, etc.). If configured → run the fetch command to get objective data.

If no integrations → skip this step.

4. Conversational Intake

Don't present a form. Have a natural conversation.

Ask open-ended: "How was your day?" or "What happened today?"

Based on what user shares, ask follow-up questions:

• If they mention symptoms → ask severity (1-10)
• If they mention exercise → ask details (type, duration, intensity)
• If they mention food → note it
• If they mention medication changes → note and flag for stack.md update
• If they mention mood/energy → capture the rating

ALWAYS ask for subjective metrics that are in the YAML template:

• Mood (1-10)
• Energy (1-10)
• Any condition-specific subjective metrics from their template

NEVER guess subjective data. If user doesn't answer → leave as null.

5. Write the Log

YAML frontmatter:

• Fill in date
• Fill in objective metrics from wearable data (if available)
• Fill in subjective metrics from conversation
• Unknown values → null (never guess)

Notes section:

• Summarize what user shared in their own words
• Include relevant details: what they did, ate, how they felt
• Mention any medication/supplement changes

AI Observations section:

• Patterns noticed (compare to recent days)
• Anomalies (anything unusual in the data)
• Connections to previous days (e.g., "Sleep was poor — yesterday was a high-stress day")
• Connections to experiments (if active experiment, note relevant data)
• Questions for the user (if something seems off)

6. Save

Save to: logs/daily/YYYY-MM/YYYY-MM-DD.md

After saving:

• If yesterday's log needs updating (delayed data from wearable) → update it
• If user mentioned something that affects another file (stack change, new symptom) → note it and offer to update

Rules

• Date in YAML = date of the log. Data from Jan 23 goes in 2026-01-23.md.
• If data belongs to a different day → put it in that day's file.
• Subjective data = ALWAYS ASK. Never estimate mood/energy from objective data.
• If something is unknown → null, never guess.
• YAML must be accurate. Correlations depend on correct data in correct files.

If user provided an argument

/log → log today /log yesterday → log yesterday /log 2026-03-15 → log specific date