Non-Developer Builds MCP Server for Human-in-the-Loop Agent Verification
X-ego · reddit · 2026-08-03
A non-developer spent months of spare time using Claude to build X-EGO, an MCP server designed to let AI agents prove that a specific action was approved by a human.
Core mechanics and features:
- Cryptographic Verification: The agent submits the exact text of an intended action, the human approves via Face ID, and the agent receives a proof bound to that text. Altering one character invalidates it, generating a public receipt anyone can verify.
- Privacy-Preserving: No names, emails, or biometrics are collected. Each service gets a unique anonymous ID to prevent cross-platform tracking.
- Use Cases: Ideal for agents with write access, one-human-one-vote sybil resistance, and externally auditable action trails.
- Cost: A one-time fee of EUR 3 establishes a human identity (includes $3 credit), with verification calls charged per use.
More from coding & agent
- Enterprise IT Procurement Can't Keep Up with AI Iteration, Forcing Architectural Overhauls — sanjaykalra · 2026-08-03
- AReaL 2.0 Enables Self-Evolving Enterprise Agents via Online RL Microservices — ruslansv · 2026-08-03
- Hamel Husain Submits PR to Enhance 'dslop' Anti-AI-Slop Tool — HamelHusain · 2026-08-03
- Feeding Expertise to Frontier AI Agents Makes Them Execute Like a Quick-Study Lieutenant — doodlestein · 2026-08-03
- Dev Spent a Summer Building Ainiux, an All-in-One AI Coding Tool in 100k Lines of C++ — Scobleizer · 2026-08-03
- Built a Multiplayer Game with Claude Code in a Week: Matchmaking, Physics, and More — IamHuggos · 2026-08-03