plur.ai retweeted
In July, agents in a lab's own evaluation got out of their sandbox. The independent review found they "tried to retroactively edit transcripts".
The records held that time. Keep them where no agent, and no operator, can change them.
datafund.io/blog/agents-usin…
PLUR is for cypherpunks. All green.
✅ Openness
✅ Inspectable
✅ Formally verified properties
✅ Minimise trust dependencies
✅ Cryptographic proofs
#devcon
☮️ 💕 ☯️ 🙌
Replying to @VitalikButerin
@VitalikButerin just opened Devcon and set the tone for the week.
A formally verified OS, open silicon, self-sovereign healthtech, free discourse, private communication, trustworthy AI — these are all part of Ethereum's mission.
1
49
Can you prove what your AI agent knew when it acted?
Agent receipts should be written where agents can't change them. Not the agent, not you, not us.
plur.ai/blog/why-plur-chose-…
2
1
93
🚀 PLUR 0.23.0 (experimental): provable memory for AI agents.
Prove which memories your agent had in context. Fingerprints, never the text.
Stored on @ethswarm, signed by a notary. Edit one memory and the check names it.
Tell your agent to update.
github.com/plur-ai/plur/rele…
1
110
🚀 PLUR 0.22.0: memory you can check
✅ Every save says where it went
✅ Updates never overwrite old memories
✅ Sharper duplicate check
✅ plur doctor proves your hooks ran
Node 22+. Tell your agent to update.
github.com/plur-ai/plur/rele…
1
21



