HN Showcase
Back to browse
EK-1 – A local-first, sovereign AI agent built in Go and Rust

EK-1 – A local-first, sovereign AI agent built in Go and Rust

by felixche·Feb 26, 2026·2 points·0 comments
Visit ProjectView on HN

AI Analysis

●MidBold Bet

Ambitious personal AI agent on Solana, but 'shadow mode' MVP and BUSL licensing slow traction.

Strengths
  • •Three-phase trust model (shadow → hand → voice) is thoughtfully designed to build confidence gradually
  • •Values-as-Code approach to decision rules is a clever way to make agent behavior auditable and user-aligned
  • •Solana on-chain reputation ledger + zero-knowledge proofs show genuine cryptographic rigor, not hype
Weaknesses
  • •Waitlist/GitHub-only—no shipped MVP, no real users, no proof the three-phase model works in practice
  • •'Personal AI that trades Solana and negotiates with other AIs' is compelling storytelling but vastly overscopes a debut product
Category
AI/ML
Target Audience

Privacy-conscious individuals; Solana users; people seeking personal AI sovereignty (early-stage community)

Similar To

Anthropic Claude (agent architecture) · Solana-based identity projects

Post Description

I’ve spent the last few months thinking about why personal AI feels like a leash rather than a tool. Current "agents" are just thin wrappers over centralized LLMs that harvest your data to sell it back to you. You still have to do the work and make decisions whenever a tool proposes a solution. More pressure on you.

I’m building Ego-Kernel (EK-1). It’s an attempt to build a "Unitary Intelligence" - a single, local-first agent that manages your digital life, financial trades on Solana, and social reputation without your data ever leaving your enclave.

The Tech Stack:

The Brain (Go): A high-concurrency orchestrator running locally. It handles "Titan Handshakes" (P2P agent negotiations) and indexes your local life-data.

The Court (Rust/Anchor): An on-chain Reputation Ledger on Solana that makes "Social Honor" unforgeable and programmable.

The Shield (BUSL 1.1): I’m using the Business Source License to prevent the "Big Tech Vampirism" we saw in '24-'25. It’s free for individuals, but enterprises pay their share.

Why this is different: Most AI projects are trying to be the next big chatbot. I’m trying to build a protocol for cognitive independence. The system uses a "Value-Weighting Matrix" that you tune locally. If your stress (biometrics) is high or your bank balance is low, your agent's aggression threshold shifts automatically.

The Manifesto explains the "Soul Drift" problem and how we're solving it through graduated autonomy.

GitHub: https://github.com/EgoKernel/The-Protocol Manifesto: https://github.com/EgoKernel/The-Protocol/blob/main/Manifest...

Would love to hear your thoughts on the local-first approach and the staking-for-reputation model.

Similar Projects

AI/ML●Mid

Pi, written in Rust

The repo looks like an agent framework implemented in Rust (Cargo.toml, src, examples) with explicit Slack webhook and Claude integration notes — sensible if you want a Rust-native assistant runtime. It's low-star and the README/landing copy don't make the unique selling point clear, so buyers will need to dig into examples to judge whether it beats established Python/JS agent toolkits.

Niche GemShip It
yajatgulati
103mo ago
Developer Tools●●Solid

Assign tasks to 7 AI agents with -mentions, autonomous mode, OpenClaw

Mysti makes multi-model coding workflows tangible: you can inline-route tasks with @-mentions and have agents execute a pipeline where each one gets the previous output, plus auto-retries for failures. The OpenClaw daemon, WebSocket streaming, status-bar provider switching, and autonomous/semi-autonomous modes show this is more than a toy — it aims to make cross-model review and debate a practical part of your edit loop. The real test will be subscription/config friction and whether multi-agent noise actually improves real-world code quality, but the feature set is a smart, ambitious bet.

Big BrainSlick
bahaAbunojaim
123mo ago