NVIDIA NeMo Fabric

NeMo Fabric gives applications, evaluation systems, and rollout platforms a consistent interface for running agent harnesses and custom agents.

• Compose runs: Configure models, tools, skills, MCP servers, subagents, and runtime settings consistently.
• Validate before execution: Detect unsupported configurations before starting a runtime.
• Switch harnesses: Reuse the consumer integration while adapters handle harness-specific behavior.
• Compare outcomes: Collect normalized results, errors, artifacts, lifecycle events, and telemetry references.

Supported Agent Harnesses

NeMo Fabric includes adapters for the following harnesses and execution targets. Select a harness name for its adapter documentation:

Agent Harness	Tool Policy	MCP	Skills	Telemetry
Claude Code	✅	✅	✅	✅
Cline	✅	✅	✅	—
Codex	—	✅	✅	✅
Factory Droid	✅	✅	✅	—
Hermes Agent	✅	✅	✅	✅
Kilo Code	✅	✅	✅	—
LangChain Deep Agents	✅	✅	✅	✅
mini-SWE-agent	—	—	—	✅
NOOA	—	✅	✅	✅
OpenClaw	✅	✅	✅	—
OpenHands	✅	✅	✅	—
OpenCode	—	✅	✅	—
Pi	✅	—	✅	✅
Qwen Code	✅	✅	✅	—
Remote Agent	—	—	—	✅

✅ Supported · — Not supported. Support may come through normalized configuration, native harness behavior, or a specific target. Refer to each adapter guide or the complete compatibility matrix for details and limitations.

NVIDIA Ecosystem

NeMo Fabric is an execution layer for NVIDIA products that build, evaluate, deploy, and improve agents. The ecosystem includes shipped integrations with NVIDIA NeMo Helix, concrete integrations with NVIDIA NemoClaw and NVIDIA NeMo Gym, and emerging rollout integrations such as NVIDIA Polar. Explore the NeMo Fabric ecosystem.

Extensibility

NeMo Fabric separates consumers from execution targets through a stable, versioned contract. This makes it extensible in two directions:

• Integrate a consumer: Applications, evaluation systems, and rollout platforms use the Python SDK and typed FabricConfig to run different harnesses without implementing each harness integration. Start with the consumer integration overview or consumer integration skills.
• Add an execution target: Adapters translate the NeMo Fabric contract into the native behavior of a harness, framework, custom agent, or remote service. Start with the adapter contract and adapter examples.

Documentation

• Quickstart: Run a first agent through the Python SDK.
• Conversation telemetry: Use RunRequest.relay_session_root to group Deep Agents and mini-SWE-agent turns under one NeMo Relay session.
• Installation: Install the runtime, adapters, harness dependencies, and telemetry components, and choose a deployment scenario.
• Adapter catalog: Inspect bundled adapter metadata on a host without installing harness SDKs.
• Host admission: Validate requested capabilities and reject host/task descriptor drift before execution.
• NeMo Fabric runtime server: Control one SDK runtime inside one sandbox through fabric-agent-ctl and one private local socket. Unconfirmed cleanup retains runtime ownership and blocks replacement starts.
• Documentation overview: Understand the execution model and choose an interface.
• Adapter reference: Compare adapter configuration, runtime, and observability support.
• Contribution guide: Build, test, and contribute to the project.

Roadmap

Planned areas of development include:

• Harness coverage: Expand supported harnesses and adapter capabilities.
• Configuration explainability: Explain native configuration, mappings, precedence, and unsupported fields; redact secrets without launching harnesses or modifying workspaces.
• ACP integration: Support negotiated sessions, streaming, cancellation, permissions, and capability-dependent resume through Agent Client Protocol.
• Sandbox execution: Support additional sandbox providers, configurable access boundaries, provider-supported sudo, and judge-code isolation.
• Environment bootstate: Capture files, configuration, dependencies, and supported image snapshots for restoration and branching evaluations.
• System statistics: Correlate CPU, memory, elapsed time, available GPU usage, and available token/cost metrics with Fabric execution IDs; identify unavailable measurements.