Seal the context
Encrypted inputs enter a protected compute environment. No raw data leaves the boundary.
Experimental protocol / 001
PRIVATE
INTELLIGENCE
FOR AN OPEN WORLD
A privacy layer for intelligent systems.
Compute beyond observation.
LOCAL / PRIVATE∅ 2025
01The idea
Intelligence should be
private by default.
Today, the most capable systems demand a trade: share your context, or stay out of the loop. TENEBRIS removes that trade.
We are exploring a new category of infrastructure where autonomous agents can reason, coordinate, and act on sensitive information — without ever revealing it.
“
Privacy is not the absence of information.
It is control over who can act on it.
THE COORDINATION LAYER$TENE
02The token
Value for the
unobservable.
$TENE is the native coordination asset for the TENEBRIS protocol — aligning private compute, agent operators, and the open network.
ASSET$TENE
ROLENative coordination asset
CHAINSolana
STATUSDEPLOYED
Contract address2F12EQa7bxc9RpPNvyyaCWkJXfqy79NBprXmsmbmpump
03Architecture
A new primitive
for agency.
/ HOW IT MOVES01
Private Compute
The environment stays sealed.
→02
Encrypted Data
Context remains yours.
→03
AI Agents
Systems reason privately.
→04
Solana
Outcomes settle openly.
04Protocol
Make the invisible
useful.
TENEBRIS is a set of cryptographic and computational primitives for agents that need to operate in the real world without making private inputs public.
Read the protocol briefReason in shadow
Agents act on the signal, not the source. Outputs are selectively revealed and verifiable.
Prove the outcome
Solana records what matters: coordination, incentives, and the final state.
05Roadmap
From research
to reality.
/ 2025—202701COMPLETE
Research
Foundational cryptography & private inference
02IN PROGRESS
Prototype
First agent runtime in a sealed environment
03NEXT
Testnet
Open network for private agent coordination
04FUTURE
Network
Permissionless intelligence layer
06Research log
Signals from
the frontier.
/ FIELD NOTES 001—003Private inference is a systems problem
Model weights are only one part of the boundary. The runtime, memory, and output all need a way to stay selectively legible.
Read noteAgents need a right to opacity
Autonomous systems cannot be useful if every action requires revealing the context that made it possible.
Read noteSettlement does not require exposure
Open coordination and private computation are not opposites. They are layers of the same network.
Read note07The thesis
Privacy is
infrastructure.
When intelligence becomes ambient, privacy can no longer be a setting. It has to be a property of the system itself.
TENEBRIS is building toward that system.
08FAQ
Questions in
the dark.
TENEBRIS is a research direction for private intelligence: a way for AI agents to act on sensitive context without exposing the context itself.
09Community
Follow the signal.
Research notes, launch updates and the $TENE contract announcement — first on X.