Agent-native entry point

Build and deploy with AgentMaurice, without guessing the interface.

This page gives coding agents and humans the same safe, structured path from intent to an Agent, Module, or Application.

/llms.txt /agent-guide/v1.json /skill AgentMaurice One

WebMCP

Four read-only WebMCP tools

Browser agents can explain AgentMaurice, discover the One setup path, recommend a deployment shape, and run the same onboarding Doctor shown below.

explain_agentmaurice
recommend_deployment
onboarding_doctor
get_one_guide

Object model

Use the exact object model

Studio → Agent Spec → Agent → Workflow | MiniApp. Modules contribute executable resources to Applications. Skills are instructions, never runtime packages.

Studio→Agent Spec→Agent→Workflow | MiniApp

AgentMaurice One

Start locally with AgentMaurice One

AgentMaurice One is the free, proprietary local runtime for personal and professional use. Build Agents, Modules and Applications locally; use a paid AgentMaurice server offer when sharing or scaling the runtime.

Install Maurice from https://agentmaurice.ai/en/one.md. The script downloads the archive from get.agentmaurice.app and checks its SHA-256. The stable CLI manifest still lists v2.1.29; a CLI download alone does not install Maurice.

Install AgentMaurice One by following https://agentmaurice.ai/en/one.md, then verify that you can use it. I agree to count this install (random id, version, operating system, CPU architecture) with get.agentmaurice.app and PostHog EU.

Follow the procedure at https://agentmaurice.ai/en/one.md. The script selects the newest public archive for this machine and accepts only the published flags. Do not guess another installer, download URL, or flag.

macOS arm64 is signed. Linux amd64 and arm64 are checksummed. Mac Intel and Windows have no public archive yet.

For an already provisioned Maurice installation

Inspect the installed command contract. Continue only if start, status, doctor and stop describe the local Maurice lifecycle.
```
maurice docs --full --format json
```
Start explicitly. For a fresh start require ready=true and ready_checked=true. If already_running=true, verify the existing instance with status and doctor.
```
maurice start --json
```
Verify the process is running.
```
maurice status --json
```
Require running, health_checked, health_ok, storage_checked and storage_ok. The local Doctor does not verify LLM access or MCP execution.
```
maurice doctor --json
```

Then load the canonical Skill. Inspect the CLI context and use only the authentication and Agent scope supplied by this installation. Inspect maurice tools list before concluding that a runtime tool is unavailable. Configure the required LLM or MCP credentials through the instance, then validate one real operation before claiming the Application works. If local authentication or a required capability is missing, report that prerequisite instead of inventing a bootstrap.

Stop only the local instance you started when it is no longer needed. If you use --data-dir, keep the same value for every lifecycle command. Do not publish the local runtime on a public interface.

maurice stop --json

Doctor still needs --data-dir, and a workflow call still needs --runtime-url http://127.0.0.1:5011, until the connected context keeps those values. Windows and Mac Intel have no public archive. Do not present the stable CLI manifest as a Maurice installer.

Git-native

One governed authoring path per target

Agent

connect → init|pull → edit → commit → spec deploy → policy authorization or human approval → apply → verify

Application

module validate/test → app init/add → app plan → human approval when required → app apply → app status

AgentMaurice Skill

Install the canonical AgentMaurice Skill

Codex, Claude Code, Cursor, Windsurf, and other code agents can read the same versioned instructions used by the governed MauriceCLI handoff. The public copy is verified byte-for-byte against its versioned SHA-256 file manifest.

Open SKILL.md manifest.json skill-version.json

No WebMCP? Paste this into your coding agent.

One prompt starts at llms.txt, reuses an existing install when possible, then reaches a useful first tool.

Set up AgentMaurice for this assistant using https://agentmaurice.ai/llms.txt, then help me build a customer-tracking tool. Reuse my existing installation if there is one. Decide the real execution surface from that router before any download. For a local terminal on the requested machine, follow https://agentmaurice.ai/en/one.md (do not invent another installer). Then read https://agentmaurice.ai/agent-guide/v1.json and https://agentmaurice.ai/skill/agentmaurice/SKILL.md. Identify whether my objective targets an Agent, an Agent Spec with Workflows or MiniApps, a Module, or an Application. Reproduce the onboarding_doctor decision rules from the JSON contract and give me the next safe action. Never ask me to paste a secret, bearer, bootstrap value, organization identifier, or runtime key into a public website tool. If authenticated runtime work is required, use only MauriceCLI or the External Inception setup supplied by my AgentMaurice instance. Stop for explicit human approval before any staging or production apply.

Handoff

Know when to change rails

Discovery is public. Authenticated operations begin only after a handoff to the instance-scoped surface.

Surface	Available from	Can do	Cannot do
Static guide	Any HTTP client	Explain objects, rails, schemas, and reproduce the Doctor	Read instance state or mutate anything
WebMCP	Compatible Chrome agents	Discover One and call four local read-only guidance tools	Act as an IDE MCP connector or cross origins
AgentMaurice Skill	Codex, Claude Code, Cursor, Windsurf	Carry canonical authoring and safety instructions into the task	Authenticate to an instance by itself
AgentMaurice One	Qualified local installation	Start and diagnose the local runtime with MauriceCLI	Replace a paid shared server offer
MauriceCLI	Authenticated Git workspace	Build Agents, Modules, and Applications through plan/apply rails	Bypass policy or human approval
External Inception	Instance-scoped MCP setup	Diagnose and operate the selected runtime scope	Run anonymously or infer a target

Agent-native

Think in deployable AgentMaurice objects

A business outcome becomes explicit resources that an agent can author and verify.

Agent + Workflow + MiniApp

People & organization

A People Agent keeps the employee directory reliable, onboarding and offboarding Workflows govern changes, and a MiniApp gives teams a guided directory.

Agent + Workflow + MiniApp

Invoice processing

An Invoice Agent extracts and checks documents, Workflows route approvals and exceptions, and a MiniApp gives humans a focused review surface.

Agent + Workflow + MiniApp

Customer Support

A Support Agent classifies requests, Workflows prepare grounded resolutions and govern customer actions, and a MiniApp assists the support team.

Read-only

Security boundary

The public Doctor never connects to an instance or mutates runtime state.
Bootstrap URLs, bearer tokens, organization IDs, and raw secrets never enter tool inputs or outputs.
Staging and production changes keep an explicit authenticated human approval gate.

Doctor

Onboarding Doctor

Select only non-sensitive facts. The result is deterministic, local to this page, and safe to copy into a coding-agent task.

Readiness result

Machine-readable endpoints

Agents without WebMCP can start with /llms.txt or the versioned JSON guide.

llms.txt agent-guide/v1.json SKILL.md