The Problem Nobody Talks About
Bittensor has 128 subnets. Subnet owners automatically receive 18% of their subnet's daily emissions — whether they build anything or not. Some owners registered a slot, did nothing, and walked away with thousands of TAO in profit.
When you stake into one of these subnets, your TAO flows to an owner who isn't creating value. Your alpha token slowly bleeds out as emissions dilute without product traction to offset them.
Until now, there was no way to tell the difference at a glance. You had to manually check GitHub repos, track owner wallets, and monitor miner counts across 128 subnets. Nobody does that.
So we built it.
Introducing the Subnet Health Scanner
We scored every Bittensor subnet on 7 on-chain and off-chain factors, updated every 15 minutes:
| Factor | Weight | What It Measures |
|---|---|---|
| Miner Participation | 25% | Active miners, earning miners, fill rate |
| Price Trajectory | 20% | 7-day and 30-day alpha price changes |
| Root Proportion | 15% | % of validators auto-selling (high = no confidence) |
| Development Activity | 15% | GitHub commits, stars, authors, days since last push |
| Liquidity Depth | 10% | TAO in the pool — low = illiquid death spiral |
| Trade Activity | 10% | 24h buy/sell volume and ratio |
| Age Factor | 5% | New subnets get a brief benefit of the doubt |
The Results: Most Subnets Are Weak
Out of 128 scored subnets:
- 8 UNPRODUCTIVE — minimal activity, likely extracting value without contributing
- 70 WEAK — low participation or declining metrics
- 50 MODERATE — some activity but not fully productive
- 0 HEALTHY — the bar is high for a reason
That means roughly 60% of subnets show signs of weakness. If you're staking blindly, the odds are against you.
Owner Wallet Tracking: Who's Draining?
We tracked all 128 subnet owner coldkeys using the TaoStats transfer API. The results:
- 44 owners are DRAINING — more TAO flowing out than in
- 54 owners are INVESTING — adding capital back into their subnet
- 14 are INACTIVE — no recent transfers at all
An owner who is consistently transferring TAO out of their wallet after collecting emissions is a red flag. It means they're treating the subnet as a cash cow, not a product.
GitHub Activity: Who's Actually Coding?
We scanned 107 subnet GitHub repositories for real development activity in the last 30 days:
Most Active (30-day commits)
| Subnet | Commits | Authors | Stars |
|---|---|---|---|
| SN62 Ridges | 100+ | 6 | 84 |
| SN114 SOMA | 100+ | 10 | 7 |
| SN56 Gradients | 98 | 4 | 34 |
| SN100 BASE | 87 | 2 | 160 |
| SN67 Harnyx | 90 | 3 | 4 |
Dead Repos (GitHub exists, 0 commits in 30 days)
| Subnet | Last Push | Stars |
|---|---|---|
| SN10 Swap | 230 days ago | 33 |
| SN43 Graphite | 254 days ago | 15 |
| SN52 Dojo | 232 days ago | 0 |
| SN25 Mainframe | 267 days ago | 30 |
Some of these subnets have 30+ stars and were once promising projects. But zero commits in 7-9 months while still collecting owner emissions? That's the definition of extracting without building.
Red Flags to Watch For
The health scanner flags specific warning signals on each subnet:
- Zero miners — nobody is doing the work
- Unnamed — subnet hasn't even been named
- Validators dumping — root proportion above 50%
- Price crash — sustained alpha decline
- GitHub stale — repo exists but no commits in months
- Owner draining — frequent outgoing transfers from owner wallet
How to Use This
Before you stake TAO into any subnet:
- Check the Subnet Health Scanner — is it MODERATE or above?
- Look at the Owner Wallet column — INVESTING is good, DRAINING is bad
- Check GitHub 30d — any commits at all?
- Look for red flags — more than 2 is a strong avoid signal
The health scanner updates every 15 minutes with real on-chain data, GitHub API data, and TaoStats wallet tracking. No manual research needed.