⌕
Discovery
Making agents easier to find across indexes and registries.
AGENTIC SYSTEMS / 01
Building
AI agents.
I’ve been exploring what happens when AI agents stop working alone.
A small space for the systems that make agents easier to discover, connect, and build with.
↓ SCROLL TO EXPLORE✦✦Same humans.
Bigger possibilities.
Bigger possibilities.
╱
╱
╱
╱
╱
CONNECTING
AI AGENTS
EVERYWHERE.
AI AGENTS
EVERYWHERE.
DISCOVER
↳ VERIFY
↳ COLLABORATE
↳ BUILD
↳ VERIFY
↳ COLLABORATE
↳ BUILD
WHY THIS INTERESTS ME / 02
The missing
infrastructure.
Most AI agents still live in silos. I’m interested in the public infrastructure that helps agents across ecosystems discover one another, exchange capabilities, and work together reliably.
different tools.
same goal.
same goal.
WHERE I CAN CONTRIBUTE / 03
↔
Interoperability
Helping different protocols and agent frameworks work together.
◉
Reliability
Exploring trust, capability evidence, and evaluation.
⌘
Developer Experience
Better tools, tests, docs, and examples for builders.
IDEAS I’D LOVE TO EXPLORE / 04
agent.trace ↗
Visual debugging for request → index → registry → runtime.
› searching agents...› index matched ✓› agent card loaded ✓› runtime reached ✓› doneprove.it ↗
Attach evidence to capability claims, not just descriptions.
claims.
backed
by proof.
backed
by proof.
Confidence-aware discovery
Let agents express how sure they are, not just yes/no.
Document-agent network
Independent extraction, tamper-check, and validation agents working together.
WHAT I BRING / 05
- Python / FastAPI / APIs
- TypeScript / Next.js / UI tools
- PostgreSQL / data systems
- LLM + RAG + agent systems
- ML evaluation / calibration / experiments
- Document intelligence
HOW I’D START / 06
READ
RUN
SHIP
WHAT I WANT TO BUILD / 07
Agents should work together.
for real.
Discovery, interoperability, reliability, and evaluation are the places I’m most curious about. Give me a system to understand, something useful to build, and people I can learn and work with.