Skip to content
Documentation
Build, run, and govern AI agent systems
The platform documentation covers the workspace, the Knowledge Hub, teams, agents, skills and tools, with the prebuilt teams documented alongside them. The developer reference is being published in parallel.
Start here
Live Getting startedWhat Nagent is, the surfaces it provides, the five core concepts behind it, and the four stages that take you from sign-up to a running agent. Read the guide→ Live The Knowledge HubHow Nagent stores organisation knowledge across enterprise, team and agent levels, and how semantic and direct retrieval serve it to agents. Read the guide→ Live Tools and integrationsHow agents connect to third-party tools, why every action is whitelisted and scoped individually, and how triggers let events start agent work. Read the guide→Teams
Live Setting up the Team WorkbenchWhat an AI Team space holds, the two ways to set one up, and why trust stays with each agent even once it is working inside a team. Read the guide→ Live Anatomy of a teamWhat a team space is made of on Nagent: the shared thread, the file record behind it, and how human and agent members hold the same standing. Read the guide→ Live The team filesHow a team's charter, roster, memory, tasks, decisions, access and events log work as readable files that both humans and agents cite. Read the guide→ Live Extending a teamHow to add people and agents to an existing team, designate leads, remove members, and grow a team without flooding the approval queue. Read the guide→Agents
Live What is an agentHow Nagent defines an agent as a governed digital worker, when a custom agent is worth building, and how the control plane differs from builder tools. Read the guide→ Live Creating an agentThe three-step process for building a new agent on Nagent: open the Agent Workbench, describe the job, then review and provision the agent. Read the guide→ Live Agent WorkbenchHow to read the agent register, what each column and agent state means, how agents group into clusters, and where autonomy and model are set. Read the guide→ Live SkillsWhat a skill is on Nagent, the fields and markdown body that make one up, and how skills are created, scoped to agents, and versioned. Read the guide→Prebuilt teams
Live Prebuilt teamsThe levels a Nagent organisation is arranged in, the prebuilt departments available, and which agents in each are live or still coming. Read the guide→ Live NORA · outbound research and activationHow NORA sources accounts, researches them into cited dossiers, sequences outreach across three channels, and holds every send behind a human approval. Read the guide→ Live Running outbound campaigns with NORAHow to build, save, schedule and monitor outbound prospecting campaigns in NORA, and turn a chat-generated ICP into leads and outreach. Read the guide→ Featured guides
Start here
End-to-end how-to guides. The multi-agent workflow guide is the first published; running individual agents and best practices are next on the queue. Complete guide: building a multi-agent workflow system The end-to-end how-to. From a one-sentence Helix brief to a governed, observable, self-improving multi-agent system on the substrate. 8-step build pathHelix synthesis to deploymentGovernance + budget capsSandbox evals + Karmic feedback Read the guide→ Running individual agents Deploy a single agent: configure guardrails, hook triggers, observe runs, and iterate. For when you don't need a full multi-agent workflow. Agent lifecycleTriggers + guardrailsObservability tab walkthroughIteration patterns Notify me→ Best practices for agentic systems Field-tested patterns for autonomy tuning, oversight design, prompt versioning, eval suites, and cost governance from production deployments. Autonomy ladder tuningPer-action override designEval-driven skill versioningCost governance patterns Notify me→
Still being written
Deep dives in progress
Reference documentation for the Agent Control Plane, Workbench, Topology, Smriti, Karmic, Helix, BuildCraft, and the governance substrate. While these are being written, the Technology Overview covers every layer in narrative form. Coming soon Agent Control PlaneThe live operator surface: SSE stream of every action, decisions ledger replay, kill switch, materialised view at fleet scale. Notify me→ Coming soon TopologyLiving-systems visualisation. Cluster agents into SYSTEMs, see cognitive gravity, matrix view, time-everywhere replay. Notify me→ Coming soon Smriti · memory fabricPersistent per-entity memory. Cross-agent + cross-thread continuity, structured sections, continuous synthesis. Notify me→ Coming soon Karmic · outcome feedbackOutcome-linked attribution, per-agent trust scores, auto-demote on drift, template + prompt promotion. Notify me→ Coming soon Governance + autonomy ladderL0 to L4 autonomy, per-action overrides, kill switch, multi-party approvals, budget caps, JIT access, MFA. Notify me→ Coming soon Helix · Agent StudioChat-first agent + multi-agent-workflow synthesis. Builder UX, AgentSpec contract, auto-fixer, eval attachment. Notify me→ Coming soon BuildCraft · pro-code SDKTypeScript + Python SDK with Git-native workflow. 9 modules (Spec · Tools · Skills · Triggers · Data/RAG · Governance · Observability · UI · Deployment). Notify me→
Developer reference
API + SDK + integration surfaces
The traditional reference docs: request/response shapes, error codes, webhook signatures, auth flows. Generated from the live OpenAPI spec, publishing shortly. QuickstartComing soon 10-minute walkthrough: API key to first webhook delivery. Notify me→ Agent SDKComing soon Python + TypeScript SDKs, type defs, worked examples. Notify me→ REST API referenceComing soon Endpoint reference, request/response shapes, rate limits. Notify me→ WebhooksComing soon Real-time event subscriptions + signature verification. Notify me→ AuthenticationComing soon API keys, OAuth, SAML SSO, rate limits, rotation. Notify me→ ConceptsComing soon How Nagent works under the hood: substrate primitives. Notify me→
Need help building?
Our solutions team will pair-program with yours to get the first agent live in days, not weeks, and walk through the 8-step path on your actual workflow.
Talk to a solutions engineerStay updated on AI: subscribe to our newsletter.