fh-lab-results

name	fh-lab-results
description	Query Function Health lab results, check for new data, compare visits, and deep-dive on biomarkers. Use when the user asks about labs, bloodwork, biomarkers, or health markers.
argument-hint	[summary | check | sync | biomarker <name> | changes | out-of-range | category <name>]
allowed-tools	mcp__function-health__fh_login, mcp__function-health__fh_status, mcp__function-health__fh_results, mcp__function-health__fh_biomarker, mcp__function-health__fh_summary, mcp__function-health__fh_categories, mcp__function-health__fh_changes, mcp__function-health__fh_sync, mcp__function-health__fh_check, mcp__function-health__fh_recommendations, mcp__function-health__fh_report, mcp__function-health__fh_version, mcp__function-health__fh_notifications

Lab Results Skill

Query and analyze Function Health lab results. This skill wraps 13 MCP tools for a conversational lab results experience.

First-Time Setup (Onboarding)

Always start by calling fh_status to check if the user is authenticated and has data.

If authenticated is false:

1. Tell the user they need to connect their Function Health account
2. Instruct them to run npx -y -p function-health-mcp function-health login in their terminal (this keeps their password secure with hidden input)
3. Once they confirm they've logged in, run fh_sync to pull their data
4. Then proceed with the requested action (or show a summary)

If authenticated is true but hasData is false:

1. Run fh_sync to pull data
2. Then proceed with the requested action

If tokenValid is false (expired session):

1. Check the authHint field in the status response for guidance
2. If FH_EMAIL and FH_PASSWORD env vars are set, auto-login will be attempted on the next API call — just retry the operation
3. Otherwise, instruct the user to run npx -y -p function-health-mcp function-health login in their terminal
4. Mention they can set FH_EMAIL and FH_PASSWORD environment variables for hands-free re-authentication

When to use

Trigger when the user mentions:

• Lab results, bloodwork, biomarkers, health markers
• Specific biomarker names (e.g., "vitamin D", "testosterone", "A1C")
• Checking for new results or syncing data
• Comparing visits or tracking changes
• Health summary or out-of-range markers

Arguments

Parse from $ARGUMENTS:

Argument	Action
summary or no args	Health summary (default)
check	Lightweight check for new results
sync	Pull latest data from Function Health
biomarker <name>	Deep dive on a specific biomarker
changes	Compare last two test rounds
out-of-range	Show all out-of-range markers
category <name>	Show results for a category (e.g., "heart", "thyroid")
recommendations	Show health recommendations
report	Show clinician report

Workflows

Default: Health Summary

1. Call fh_summary
2. Present conversationally:
   • Total markers tested, how many in/out of range
   • Biological age vs chronological age (if available)
   • BMI (if available)
   • List out-of-range markers with values
3. If there are out-of-range markers, offer to deep-dive on any of them

Check for New Results

1. Call fh_check
2. Report pending/completed requisitions and last sync time
3. If newResultsAvailable is true, ask if user wants to sync

Sync

1. Call fh_sync
2. Report how many results were pulled and if there are new ones
3. If new results found, automatically run a changes comparison

Biomarker Deep Dive

1. Call fh_biomarker with the name
2. Present:
   • Current value and whether it's in range
   • Optimal range
   • History across all visits (show trend direction)
   • Why it matters (clinical context)
   • Recommendations (foods, supplements, lifestyle)
3. Keep it conversational — lead with the number, then context

Daily Briefing

For daily review / morning briefing, include a Function Health section:

1. Call fh_notifications (without acknowledge) to check for pending changes
2. If no notifications → report "No new lab data" and move on
3. If there are notifications, present them:
   • Lead with what's significant: worsened markers, biological age changes
   • Then improvements and new results
   • For new results, automatically call fh_changes to compare against previous rounds of the same tests — show how values moved
   • For any newly out-of-range marker, call fh_biomarker to get context and recommendations
4. After presenting everything, call fh_notifications with acknowledge: true to clear

This ensures the user gets a complete picture without needing to ask follow-up questions.

Example output:

Function Health: New lab results came in yesterday. 3 new results — Vitamin D improved to 42 (was 28, now in range), A1C steady at 5.4. LDL moved out of range at 142 (was 128). I'd recommend focusing on the LDL — here are the dietary recommendations...

Compare Test Rounds

1. Call fh_changes
2. Each test round includes all lab visits (1-3) over several weeks. Comparisons are between complete rounds, not individual visit dates.
3. Summarize by category:
   • Improved markers (moved into range)
   • Worsened markers (moved out of range)
   • Significantly changed (>10% delta)
   • New markers (first time tested)
4. Lead with wins, then concerns

Out-of-Range Review

1. Call fh_results with status: "out_of_range"
2. Group by severity or category if many results
3. For each, show name and value
4. Offer to deep-dive on specific markers

Category View

1. Call fh_results with the category name
2. Show all markers in that category with values and status
3. Highlight any out-of-range

Presentation Style

• Conversational, not clinical. "Your Vitamin D is at 28 — a bit below optimal (40-60). It's been trending up though, from 22 last visit."
• Lead with what matters. Out-of-range first, then context.
• Use specific numbers. "72 out of 78 markers in range" not "most markers look good."
• Show trends when available. "Up from 145 → 162 since last visit."
• Offer next steps. After showing out-of-range, offer to deep-dive. After showing a biomarker, mention if there are recommendations.

Periodic Monitoring

Change notifications persist across conversations — missed days don't lose data. New lab results from Function Health typically arrive in batches over several weeks as different panels complete. The daily briefing workflow (above) is one way to surface these; alternatively, just run /fh-lab-results check periodically.