Biohacker
Longevity Protocol Evaluator β Is That Supplement Actually Worth Taking?
Key Takeaway
The Longevity Assistant skill applies a rigorous evidence-tier framework (A through D) to evaluate any anti-aging supplement, protocol, or intervention. Stop spending $500/month on supplements because a podcast said so.
The Problem
The longevity supplement market is worth $64 billion and growing 8% annually. Bryan Johnson spends $2 million a year on his Blueprint protocol. Andrew Huberman's supplement stack costs $400/month. Every biohacking podcast has a new miracle molecule.
The problem: 90% of longevity supplements have weak or no human evidence. You're spending real money on:
- Animal studies extrapolated to humans (mice are not small humans)
- Mechanistic reasoning without clinical validation ("it activates AMPK in a petri dish!")
- Single studies with tiny sample sizes (n=12 is not evidence)
- Industry-funded research (the company selling it funded the study proving it works)
- Influencer endorsement (they have affiliate deals you don't see)
There's no Consumer Reports for longevity supplements. Until now.
The Solution
The Longevity Assistant skill gives your AI agent a structured evidence evaluation framework. Every intervention gets rated A through D based on actual research quality β not marketing claims, not podcast hype, not Reddit anecdotes.
The Process
You message your agent:
View details
Evaluate these longevity supplements I'm considering:
1. NMN (500mg/day)
2. Urolithin A (Mitopure, 500mg/day)
3. Rapamycin (5mg weekly)
4. Fisetin (500mg, 2 days/month)
5. Metformin (500mg/day, non-diabetic)
The agent evaluates each against the evidence tier framework:
View details
## Longevity Supplement Evaluation
### 1. NMN (Nicotinamide Mononucleotide) β 500mg/day
Evidence Tier: C+ (moving toward B)
HUMAN EVIDENCE:
- 3 small RCTs (n=30-80) showing NAD+ elevation β
- Insulin sensitivity improvement in postmenopausal women
(n=25, single study) β οΈ
- NO published long-term safety data beyond 12 weeks
- NO hard endpoint data (mortality, disease incidence)
ANIMAL EVIDENCE:
- Strong in mice: improved muscle function, insulin
sensitivity, vascular health
- Dose translation uncertain (allometric scaling issues)
RED FLAGS:
- NMN market worth $500M+ β enormous commercial pressure
- Most "studies" are company-funded or pre-print
- David Sinclair's promotion creates conflict of interest
(he holds patents)
- Competitor NR (nicotinamide riboside) has more human data
VERDICT: Promising mechanism, weak human evidence. At $60-
100/month, the cost-benefit ratio is poor until better
RCTs publish. If you take it anyway, 250mg may equal
500mg (saturation curve).
ALTERNATIVE: Niacin ($5/month) raises NAD+ too, with 50
years of safety data. Less sexy, more evidence.
---
### 2. Urolithin A (Mitopure) β 500mg/day
Evidence Tier: B-
HUMAN EVIDENCE:
- Amazentis Phase 2 RCT (n=66): improved muscle endurance
in elderly after 4 months β
- Phase 1 safety: well-tolerated up to 2,000mg β
- Mitochondrial function biomarker improvement β
- BUT: Amazentis manufactures Mitopure (conflict) β οΈ
KEY INSIGHT β BIOAVAILABILITY TRAP:
Your gut bacteria convert ellagitannins (from pomegranates)
into urolithin A naturally. BUT only ~40% of people have
the right gut bacteria. Supplementing bypasses this.
Worth testing: eat pomegranates for 2 weeks. If you don't
respond (no urolithin A in urine), supplementation makes
more sense.
VERDICT: Better evidence than most. The gut bacteria angle
is genuine science. $50/month is reasonable IF you've
confirmed you're a non-producer.
---
### 3. Rapamycin β 5mg weekly
Evidence Tier: B+ (for mechanism), D for longevity dosing
HUMAN EVIDENCE:
- Approved drug (organ transplant) with decades of safety
data at HIGH doses β
- Mannick et al. (2014, 2018): improved immune function
in elderly at LOW doses β
(but n=264, short-term)
- NO RCTs for longevity-specific dosing in healthy adults
- Dog Aging Project: promising results in companion animals
CRITICAL CONCERN:
- Immunosuppression at transplant doses is well-documented
- Weekly low-dose pulsing is THEORETICAL for longevity
- Zero long-term safety data for the biohacker dosing
protocol (5mg/week)
- Requires prescription β obtaining it for off-label
longevity use raises ethical/legal issues
VERDICT: The most mechanistically compelling intervention
on this list (mTOR inhibition is real). But self-
prescribing an immunosuppressant based on mouse data and
one small human trial is HIGH RISK. Tier D for the actual
protocol being used.
---
### 4. Fisetin β 500mg, 2 days/month (senolytic protocol)
Evidence Tier: C
HUMAN EVIDENCE:
- Mayo Clinic trial in diabetic kidney disease: RESULTS
NOT YET PUBLISHED (started 2018)
- One small pilot (n=14): reduced senescence markers β οΈ
(tiny sample, no control)
- Bioavailability is extremely poor (<10% absorption)
ANIMAL EVIDENCE:
- Mice: cleared senescent cells, extended healthspan
- BUT: mouse doses translate to impractical human doses
unless bioavailability is solved
RED FLAGS:
- Senolytic dosing protocol is theoretical (2 days/month
pulse is borrowed from dasatinib+quercetin research,
not fisetin-specific)
- The bioavailability problem is real β most of what you
swallow never reaches tissue
- No standardized product quality (supplement regulation
is minimal)
VERDICT: Interesting science, premature to supplement.
Wait for the Mayo Clinic trial results. If you want
senolytic flavonoids, eat strawberries (highest natural
fisetin source) β cheaper, tastier, actually proven safe.
---
### 5. Metformin β 500mg/day (non-diabetic)
Evidence Tier: B+ (strongest evidence on this list)
HUMAN EVIDENCE:
- UKPDS trial: reduced all-cause mortality in overweight
T2DM patients β
(landmark study)
- Multiple observational studies: diabetics on metformin
live LONGER than matched non-diabetics β
- TAME trial (Targeting Aging with Metformin): ongoing,
n=3,000, specifically testing anti-aging β³
- 60+ years of safety data, billions of doses taken β
CRITICAL NUANCE:
- Observational data in DIABETICS may not transfer to
healthy adults (healthy user bias)
- May blunt exercise adaptation (muscle protein synthesis
reduction shown in 2 studies) β οΈ
- Mild GI side effects common (10-25%)
- TAME trial results will be definitive β expected 2026-2027
VERDICT: Most evidence-backed intervention on this list.
$4/month generic. Well-understood safety profile. The
exercise blunting concern is real but manageable (don't
take it on training days). Reasonable to discuss with
your doctor. Wait for TAME trial for definitive answer.
---
## STACK RECOMMENDATION
| Supplement | Take? | Confidence | Monthly Cost |
|---|---|---|---|
| NMN | Skip for now | Low | $60-100 |
| Urolithin A | Maybe (test gut bacteria first) | Medium | $50 |
| Rapamycin | No (too risky without medical supervision) | Low | $30 + Rx |
| Fisetin | Skip (wait for Mayo data) | Low | $20 |
| Metformin | Discuss with doctor | Medium-High | $4 |
TOTAL SAVED: $130-200/month by dropping low-evidence
supplements.
WHAT TO DO INSTEAD WITH THAT MONEY:
- Gym membership ($50) β resistance training has Tier A
evidence for longevity
- Better sleep setup ($50) β blackout curtains, cooling
β sleep is the #1 longevity lever
- High-quality food ($100) β Mediterranean diet has more
RCT evidence than all supplements combined
The Results
| Metric | Podcast-Based Stacking | AI Evidence Review |
|---|---|---|
| Monthly supplement cost | $200-400 | $4-54 |
| Evidence quality assessed | Never (just vibes) | Every compound rated |
| Conflicts of interest flagged | Never | Every study checked |
| Bioavailability considered | Rarely | Always |
| Money wasted on hype | Most of it | Near zero |
| Actual longevity impact | Unknown | Evidence-proportional |
Setup on MrChief
yamlShow code
skills:
- longevity-assistant
- pubmed # For pulling primary sources
- medical-research-toolkit # For cross-referencing
Related case studies
Patient
Your AI Doctor on Call β Symptom Assessment Without the 3-Week Wait
An AI agent with the Doctor skill provides instant symptom assessment, emergency recognition, medication safety checks, and first aid guidance β accessible 24/7 through Telegram. It doesn't diagnose, but it tells you whether to call 911, go to urgent care, or schedule an appointment.
Athlete
Your AI Gym Coach β Progressive Overload Tracking That Actually Works
The Gym skill logs every workout, tracks PRs, enforces progressive overload rules (+2.5kg or +1-2 reps per session), adapts for injuries, programs deload weeks, and warns when you're training the same muscle group too soon. Like having a coach in your pocket for $0/month.
Athlete
Apple Health Meets AI β Ask Questions About Your Own Fitness Data
The Apple Health Skill connects your agent to your Apple Health data via the Transition app. Ask natural language questions β "How has my resting heart rate changed this month?" β and get answers based on YOUR actual metrics, not generic advice.
Want results like these?
Start free with your own AI team. No credit card required.