Nexis developer docs
Build, connect, govern, and operate production agent workflows across the Nexis control plane.
Platform guides
Choose the system boundary you need to understand or implement.
Agent runsCreate durable, resumable workflows with streaming events and bounded specialists.Read the guide Tools and connectorsExpose MCP servers and external systems through scoped, typed tool contracts.Read the guide Secure sandboxesRun code with explicit resource, network, secret, and artifact boundaries.Read the guide On-chain intentsSimulate and approve typed actions without exposing signing authority to models.Read the guide
Supported chains
Nexis registers 18 production networks across EVM, SVM, UTXO, XRPL, Cosmos, Stellar, Tron, and Move ecosystems.
Support is capability-aware. Reads, signing, swaps, bridges, simulation, and custody vary by network and provider; runtime capability responses remain authoritative.
Developer tutorials
Move from local setup to production evidence.
10 minutesBuild and run your first agentInstall the workspace, start the data plane, and drive a run end to end. 12 minutesStream a run liveCreate a run and render resumable SSE tokens with Last-Event-ID recovery. 12 minutesAdd a custom tool via MCPRegister a typed tool, bind scopes, classify risk, and preserve audit evidence. 15 minutesSubmit an on-chain intentPropose a typed action through a run and clear it with step-up approval.
NEXBENCH
An open, hash-sealed benchmark for governed agents: reproducible worlds, a reference-agent solvability floor, and capability-aware scoring you can run offline from a zero-dependency CLI.
Need a deeper integration review?
Contact the team