Wispr Flow-Now on Android!
Mini tools
@alexgarden-mnemom
Founded Relic Entertainment at 22 (Homeworld, Company of Heroes). Scaled Xbox LIVE at Microsoft. Led Zynga Studios. Now building trust infrastructure for AI agents at Mnemom (mnemom.ai).
Follow
Alex Garden's tools
-
Trust scores, cryptographic proof, and risk assessment for AI agents.Open
- Trust Score™: Bond-rating-style scores (AAA–CCC) for every AI agent, computed from five weighted signals: integrity, compliance, drift, trace quality, and coherence - Zero-knowledge proofs: Every agent action gets a cryptographic attestation chain — Ed25519 signatures, SHA-256 hash chains, Merkle trees, SP1 STARK proofs - Team risk engine: CoVaR, Markowitz portfolio theory, and DebtRank applied to multi-agent fleets — know which agent is dragging quality down - Real-time containment: Pause, kill, or resume agents automatically when trust drops below threshold Smoltbot gateway: One environment variable, zero code changes — supports Anthropic, OpenAI, and Gemini - Open protocols: AAP (behavioral contracts) and AIP (runtime analysis) on npm and PyPI
Comments
On Mnemom
Alex Garden
🛠️ 1 tool
Feb 24, 2026
@Mnemom
We built Mnemom because we were shipping AI agents into production and couldn't answer a basic question: how do you prove this thing is doing what you told it to? Not monitor. Prove.
So we built the proof layer. Cryptographic attestations, trust scores, risk assessment, containment — everything you need to deploy agents you can actually stand behind. It's free to start and open source.
We'd love to hear what you think - new features shipping daily.
