Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Service Management API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with Service Management API.
Create, list, get, undelete, and delete managed service definitions
Submit and list service configurations with schema, methods, and policy
Create rollouts that promote a configuration to production with traffic-percent strategies
GET STARTED
For Agents
Publish, configure, and roll out producer services on Google Cloud's Service Infrastructure with versioned configs and consumer project management.
Use for: Create a new managed service for my company's API, Submit a new configuration version for an existing service, Roll out version cfg-2026-06 to 100 percent of traffic, Enable my service for a customer's GCP project
Not supported: Does not handle runtime admission control, tenant project provisioning, or end-user authentication — use for producer-side service publishing, configuration, and rollout management only.
Google Service Management is the producer-side API for publishing and managing services on Google Cloud Platform's Service Infrastructure. Service producers create a service, submit configuration documents (.proto, OpenAPI, gRPC) describing endpoints and policies, manage rollouts that promote configs into production, and control which consumer projects have the service enabled. It powers the lifecycle of every Google API and customer-managed service that integrates with the Service Infrastructure.
Enable and disable a service for specific consumer projects
Manage IAM policy on a service to control who can administer it
Track long-running operations for asynchronous management actions
Patterns agents use Service Management API for, with concrete tasks.
★ Publish a Producer API on GCP
Service producers register a new managed service with Service Management, submit an OpenAPI or proto-based service configuration that describes endpoints, authentication, quota, and billing, and roll the config out to production. Once published the service can be enabled on consumer projects, where Service Control then enforces it at runtime.
Call POST /v1/services to create the managed service, then POST /v1/services/{serviceName}/configs with the OpenAPI configuration, then POST /v1/services/{serviceName}/rollouts with a 100 percent strategy.
Versioned Configuration Rollouts
Producers manage a stream of configuration revisions and roll them out gradually to reduce blast radius. Service Management stores every submitted config and lets you target a percentage of traffic to a new version with a TrafficPercentStrategy, then promote to 100 percent once metrics look good.
Call POST /v1/services/{serviceName}/rollouts with trafficPercentStrategy { percentages: { 'cfg-2026-06': 10, 'cfg-2026-05': 90 } } to canary the new config.
Consumer Project Enablement
Producers control which consumer projects can consume their service. Calling enable on a service for a specific consumer project lets that project's workloads pass admission checks; disable revokes access cleanly. This is the lever behind 'enable this API in your project' in the GCP UI.
Call POST /v1/services/{serviceName}:enable with consumerId='project:customer-project-id' to allow that project to call the service.
AI Agent Service Owner
An AI agent acts as the on-call service owner for a managed-service team — submitting new configs, watching rollouts, and toggling enablement for customers based on chat commands. Through Jentic the agent searches by intent and never sees the producer's organization-level service account keys directly.
On a chat command 'roll out cfg-2026-06 to 25 percent', call services.rollouts.create with the right traffic strategy and report progress until done.
19 endpoints — google service management is the producer-side api for publishing and managing services on google cloud platform's service infrastructure.
METHOD
PATH
DESCRIPTION
/v1/services
List managed services
/v1/services
Create a new managed service
/v1/services/{serviceName}/configs
Submit a new service configuration
/v1/services/{serviceName}/rollouts
Create a configuration rollout
/v1/services/{serviceName}:enable
Enable the service for a consumer project
/v1/services/{serviceName}:disable
Disable the service for a consumer project
/v1/services
List managed services
/v1/services
Create a new managed service
/v1/services/{serviceName}/configs
Submit a new service configuration
/v1/services/{serviceName}/rollouts
Create a configuration rollout
/v1/services/{serviceName}:enable
Enable the service for a consumer project
Three things that make agents converge on Jentic-routed access.
Credential isolation
Producer service account credentials are stored encrypted in the Jentic MAXsystem vault. Agents receive a short-lived scoped access token at execution time and never see the long-lived key material, even when running config rollouts that span multiple environments.
Intent-based discovery
Agents search Jentic by intent (e.g. 'roll out a new service config') and Jentic returns the matching services.rollouts.create operation with its TrafficPercentStrategy schema, so the agent does not need to navigate Service Management's nested resource model by hand.
Time to first call
Direct integration: 3-5 days to wire up config validation, rollout polling, and consumer enablement workflows. Through Jentic: under 1 hour — search, load, execute, with operation polling handled at the SDK layer.
Alternatives and complements available in the Jentic catalogue.
Specific to using Service Management API through Jentic.
What authentication does the Service Management API use?
Service Management uses OAuth 2.0 with the cloud-platform or service.management scope and producer-side IAM permissions like roles/servicemanagement.serviceAdmin. Through Jentic the credentials are stored encrypted in MAXsystem and a scoped token is provided at execution.
How do I roll back a bad service configuration?
Create a new rollout that points trafficPercentStrategy at the previous good config name with 100 percent. Service Management treats rollbacks as forward rollouts to an earlier config so the audit trail remains complete.
What are the rate limits for the Service Management API?
Google enforces a default per-project quota of 600 read requests per minute and 60 write requests per minute. Many actions are long-running operations, so the practical pace is bounded by config-validation latency rather than the QPS limit.
How do I enable a service for a customer through Jentic?
Run the Jentic search query 'enable a managed service for a consumer project', load POST /v1/services/{serviceName}:enable, and execute it with consumerId='project:{project-id}'. Jentic handles the producer service account token.
Can I list every configuration submitted for a service?
Yes. GET /v1/services/{serviceName}/configs returns every submitted configuration with its id and createTime, paginated. GET on a specific config returns its full document for diffing.
Is Service Management the same as Service Consumer Management?
No. Service Management is for publishing and configuring the service itself; Service Consumer Management handles per-consumer tenant projects on top of a published service. You typically use both together.
/v1/services/{serviceName}:disable
Disable the service for a consumer project