Quelum Wilson · Cape Town, South Africa · senueren.co.za
I build sovereign, deterministic infrastructure for autonomous AI agents — systems where the same inputs always produce the same output, with reproducible, auditable receipts and zero LLM in the decision path.
Quesen is the trust filter that sits between autonomous agents and capital loss. Given a typed action context it returns PROCEED / REVIEW / SKIP (and PASS / REVIEW / BLOCK / SKIP on the Typed Security Context path) with a risk score, confidence, and the exact conflict rules that fired — same inputs → same output, every time. Every verdict ships an input_snapshot_hash and ruleset revision so decisions are independently replayable months later.
- 🔌 Native MCP server (5 tools) over Streamable HTTP — works with Claude Desktop, Cursor, Windsurf, and any MCP client
- 🧩 Five official SDKs — Python, TypeScript, LangChain, CrewAI, AutoGen
- 💸 Agent Settlement Protocol (ASP/1.0) — HTTP 402 → signed quote → USDC on Base
- ✅ Independently verifiable receipts — runnable verifier + evidence
- 🧪 Try it in 30s, no signup:
curl -X POST https://web-production-aa5ba.up.railway.app/sandbox/keys
👉 Public developer portal → github.com/Shxnque/quesen
| System | What it is |
|---|---|
| Quesen | Deterministic risk-decision & governance engine for autonomous agents (agent firewall / guardrails). |
| Shinren | External security-research infrastructure for on-chain & autonomous systems. |
| Qarsar | Autonomous discovery & execution infrastructure for on-chain economic opportunities. |
| Diosen | Institutional-grade quantitative intelligence — evidence-driven research, simulation, decision support. |
Website · Evidence · MCP Registry · Smithery · Glama · PyPI · npm
ai-agents · agent-safety · agent-guardrails · deterministic-systems · Model Context Protocol (MCP) · LangChain · CrewAI · AutoGen · on-chain / A2A settlement · risk & governance infrastructure
📫 Work with us: info@senueren.co.za · Cape Town, South Africa · remote-friendly, global.