INSPECTORAI-native infrastructure
for on-chain markets.
Every launch gets
inspected.
About this lab
Didge builds autonomous systems that watch on-chain markets so people don't have to.
Every market that handles real money has an inspector. Token launches never got one. Didge exists to build that layer: autonomous agents that catch a launch at block zero, keep watching, file a verdict, and seal it on-chain.
We do not rate what to buy. We make sure someone has already walked the floors.
TRUSS
The launch inspector
TRUSS is an autonomous agent that meets a token at the moment it exists and does not leave. It reads the contract, the liquidity, the holders and the deployer, runs the trade it would not ask you to run, and files one verdict. The record is sealed on-chain, so the verdict cannot be edited after the fact.
In this chapter:
- Catch it at block zero
- Keep looking: 10m, 1h, 24h, rolling
- One color: GREEN / YELLOW / ORANGE / RED
- Seal the record on-chain
- Scores are not for sale. The queue is.
02. The feed
Inspection is a stream, not a screenshot.
| TIME | TOKEN | CHAIN | T+0 | T+1H | T+24H | VERDICT |
|---|---|---|---|---|---|---|
| 14:04:52 | STRIDE | ON-CHAIN | 71 | 64 | 58 | YELLOW |
| 13:58:19 | EMBERCO | ON-CHAIN | 88 | 86 | 84 | GREEN |
| 13:41:07 | HOOFPRINT | ON-CHAIN | 44 | 29 | 21 | RED |
| 13:22:35 | LATCH | ON-CHAIN | 63 | 59 | 47 | ORANGE |
| 12:57:44 | PLUMBLINE | ON-CHAIN | 79 | 81 | 80 | GREEN |
| 12:39:02 | SIXFOLD | ON-CHAIN | 52 | 48 | 43 | ORANGE |
| 12:11:58 | DRAYHORSE | ON-CHAIN | 67 | 70 | 66 | YELLOW |
| 11:48:30 | NULLSET | ON-CHAIN | 31 | 18 | 12 | RED |
Snapshot, not advice. Scores move. The verdict is the record.
03. How a verdict is made
- 01. Ownership & mint
- 02. Liquidity lock
- 03. Holder concentration
- 04. Deployer history
- 05. Sell simulation
- 06. Rolling re-inspection
Six checks, one color, sealed on-chain.
A verdict is not an opinion about a token. It is the sum of checks that either pass or do not, weighted the same way for every launch, run again at every snapshot. These weights are the current model.
Weight in the composite score
Ownership & mint
20%
Liquidity lock
25%
Holder concentration
15%
Deployer history
20%
Sell simulation
20%
Weights are placeholder values and will move as the model is tuned.