Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Cloud Profiler 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://jentic.com/install.sh?src=apis&api=%2Fapis%2Fgoogleapis.com%2Fcloudprofiler" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fgoogleapis.com%2Fcloudprofiler" | 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 Cloud Profiler API.
Register a new continuous-profiling session for a deployment
Submit profile data captured by an in-process profiling agent
Submit offline profile data captured outside Google Cloud
GET STARTED
For Agents
Submit and patch CPU and heap profiles from applications running on or off Google Cloud. Used by profiling agents and offline tooling that feed Cloud Profiler's flame graphs and comparison views.
Use for: I need to start a CPU profiling session for my service, Submit a heap profile sample collected by my agent, Send an offline profile captured from a non-GCP host, Check whether the profiler accepted my last upload
Not supported: Does not capture metrics, run distributed traces, or store logs - use for CPU and heap profile submission to Cloud Profiler only.
Google Cloud Profiler is a managed continuous-profiling service that gathers CPU and heap profiles from production applications with negligible overhead. The API is small and primarily agent-facing: profiling agents embedded in applications call createProfile to register a profiling session, then upload sampled profile data via patch operations. Offline reporting tools use createOfflineProfile to submit profile data captured outside Google Cloud, allowing the same flame-graph and comparison tooling to surface non-GCP workloads.
Patch an active profile with sampled data and metadata
Patterns agents use Cloud Profiler API for, with concrete tasks.
★ Continuous Production CPU Profiling
Engineering teams embed the Cloud Profiler agent in production binaries; the agent calls createProfile to claim a profiling slot for the deployment, samples the running process for the configured profile type, and patches the resulting profile back to Cloud Profiler. The flame graph in the console then shows where CPU time goes across versions and zones.
Call POST /v2/{parent}/profiles with a deployment payload, then PATCH /v2/{name} with the captured profile bytes
Offline Profile Submission for Non-GCP Workloads
Teams running workloads outside Google Cloud capture profiles with their own tooling and call createOfflineProfile to submit them. The same flame-graph view and comparison tooling that supports GCP workloads is then available for on-prem or other-cloud services, giving a single profiling surface across heterogeneous deployments.
Call POST /v2/{parent}/profiles:createOffline with a populated profile message including deployment, profileType, and the captured profileBytes
AI Agent Performance Analyst via Jentic
An agent invoked during a performance investigation searches Jentic for the right Cloud Profiler operation, submits a captured profile, and points a human at the resulting flame graph. For ad-hoc analysis on non-GCP hosts, the agent uses createOfflineProfile so the same console view applies.
Search Jentic for 'submit offline profile to google cloud', execute against POST /v2/{parent}/profiles:createOffline with the captured bytes, and report the resulting profile name
4 endpoints — google cloud profiler is a managed continuous-profiling service that gathers cpu and heap profiles from production applications with negligible overhead.
METHOD
PATH
DESCRIPTION
/v2/{+parent}/profiles
Create a new profiling session for a deployment
/v2/{+parent}/profiles:createOffline
Submit an offline profile captured outside Google Cloud
/v2/{+name}
Patch a profile with captured sample bytes
/v2/{+parent}/profiles
Create a new profiling session for a deployment
/v2/{+parent}/profiles:createOffline
Submit an offline profile captured outside Google Cloud
/v2/{+name}
Patch a profile with captured sample bytes
What agents get from Jentic-routed access to this vendor.
Setup
Wiring the Cloud Profiler API by hand means setting up Google OAuth 2.0, minting short-lived tokens against the cloud-platform scope, and addressing each profile by parent resource name against cloudprofiler.googleapis.com yourself. Through Jentic you install once, import the Cloud Profiler API from the API Directory, store the OAuth credential once, and your agent calls it.
Permission scoping
Cloud Profiler puts the parent and profile resource names in the URL path (/v2/{+parent}/profiles, /v2/{+name}), so a rule can pin your agent to one project: it can create and update profiles there and nothing else. You choose the operations it may call, so the offline profile-create operation is not included unless you add it.
Credential isolation
Your Cloud Profiler OAuth credential is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
Intent-based discovery
Agents search Jentic by intent such as 'submit a CPU profile to Cloud Profiler', and Jentic returns the matching Cloud Profiler operation with its input schema so the agent calls the right endpoint without browsing the reference docs.
Alternatives and complements available in the Jentic catalogue.
Specific to using Cloud Profiler API through Jentic.
What authentication does the Cloud Profiler API use?
The Cloud Profiler API uses OAuth 2.0 with the cloud-platform or monitoring.write scope. Tokens are issued for a Google service account or end user. Through Jentic, the OAuth credential lives in your Jentic One instance and the agent receives a short-lived access token only - the underlying service account JSON never enters agent context.
Can I submit profiles from a service running outside Google Cloud?
Yes. Use POST /v2/{parent}/profiles:createOffline with a profile message that includes deployment metadata (target, labels), profileType (CPU, HEAP, WALL, etc.), and the captured profileBytes. The same console view that supports on-GCP profiling shows the resulting flame graphs.
What are the rate limits for the Cloud Profiler API?
Cloud Profiler ingestion quotas are published in the Google Cloud console under the Cloud Profiler API quota page; the API is designed for low-frequency uploads from continuous agents (typically one profile per minute per process), so per-project quotas are sized accordingly.
How do I start a profiling session through Jentic?
Search Jentic for 'create google cloud profile' and execute the returned operation against POST /v2/{parent}/profiles with a Deployment payload that names the service, version, and zone. The response includes a profile name and the agent then patches it with sampled data.
Which profile types does Cloud Profiler support?
Cloud Profiler supports CPU, WALL, HEAP, HEAP_ALLOC, CONTENTION, and THREADS profile types depending on the language runtime. The supported set is included in the createProfile request and the agent samples accordingly. Offline uploads can carry any of the supported types.
Is Cloud Profiler free?
Cloud Profiler is included with Google Cloud at no separate cost; only the runtime resources consumed by the profiling agent are billed under the host service. Pricing is documented on the Cloud Profiler pricing page.