Google Security Token Service API

Name: Security Token Service API
Brand: Security Token Service API
Availability: InStock

✓ Official Vendor SpecIdentity AuthAuthenticationoauth21 Endpoints

Install Jentic One Beta

Connect the Security Token Service API to your agent

Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Security Token Service 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 | sh

Step 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 instance

Jentic 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.

Capabilities

What an agent can do with Security Token Service API.

Exchange a third-party OIDC ID token for a Google federated access token

Mint short-lived Google Cloud tokens from AWS or Azure workload credentials

Generate scoped access tokens for impersonating service accounts via workload identity federation

GET STARTED

Start building with Security Token Service API

Explore with Jentic

View OpenAPI Document

For Agents

Exchange external (AWS, Azure, OIDC, SAML) credentials or Google credentials for a short-lived Google Cloud access token via workload identity federation.

Use for: I need to exchange an AWS IAM credential for a Google Cloud access token, Get a short-lived Google federated token from a third-party OIDC ID token, Set up workload identity federation between Azure and Google Cloud, Retrieve a downscoped access token for a specific GCS bucket

Not supported: Does not handle user authentication, password verification, or service-account creation - use for short-lived federated token exchange only.

Use Cases

Patterns agents use Security Token Service API for, with concrete tasks.

★ Workload Identity Federation from AWS

Allow workloads running on AWS EC2 or EKS to call Google Cloud APIs without distributing service-account JSON keys. The workload signs an AWS GetCallerIdentity request, posts it as a subject token to the STS token endpoint, and receives a federated Google access token scoped to an impersonated service account. This eliminates the long-lived key rotation burden and is the Google-recommended pattern for cross-cloud authentication.

POST /v1/token with grant_type=urn:ietf:params:oauth:grant-type:token-exchange, subject_token_type=urn:ietf:params:aws:token-type:aws4_request, and an AWS-signed subject token to receive a federated Google access token.

OIDC Identity Federation for CI/CD

Let GitHub Actions, GitLab CI, or any OIDC-issuing CI provider obtain Google Cloud credentials at job runtime. The CI runner presents its OIDC ID token to the STS endpoint, which validates the issuer against a workload identity pool provider and returns a Google access token. Eliminates secrets stored in CI variables and supports per-job credential scoping.

Exchange a GitHub Actions OIDC token for a Google access token by calling /v1/token with subject_token_type=urn:ietf:params:oauth:token-type:jwt and the workload identity pool audience.

Downscoped Tokens for Multi-Tenant Storage

Issue per-request access tokens that are restricted to a single Cloud Storage bucket or object prefix using Credential Access Boundary. A backend exchanges its broad service-account token for a downscoped token at the STS endpoint and hands the narrowly-scoped token to a client. Reduces blast radius when delegating storage access to untrusted code paths.

Call /v1/token with grant_type=urn:ietf:params:oauth:grant-type:token-exchange and an options field containing an accessBoundary that limits the returned token to a specific GCS bucket.

Agent Credential Brokerage via Jentic

An AI agent operating across cloud providers needs to call Google Cloud APIs but holds only AWS or OIDC credentials. Through Jentic, the agent invokes the STS token-exchange operation with its native credential, receives a short-lived Google access token, and uses it to call downstream Google APIs (Storage, BigQuery, Vertex AI). Jentic isolates the broker credential and caches the resulting token until expiry.

Through Jentic, search 'exchange aws token for google access token', load the STS token operation, and execute it with the agent's AWS subject token to obtain a federated Google credential.

Why Jentic?

Three things that make agents converge on Jentic-routed access.

Credential isolation

Your STS subject 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 'exchange an aws credential for a google token', and Jentic returns the STS token-exchange operation with its input schema, including grant_type, subject_token_type, and audience, so the agent calls the right endpoint without browsing the reference docs.

Time to first call

FAQs

Specific to using Security Token Service API through Jentic.

What authentication does the Security Token Service API use?

The STS endpoint itself is unauthenticated for the token-exchange call but requires a valid subject token (an OIDC JWT, AWS GetCallerIdentity signature, or SAML assertion) in the request body. The endpoint validates that subject token against a configured workload identity pool provider before issuing a Google access token. Through Jentic, the subject credential is held in the encrypted vault and never exposed to the agent context.

Can I exchange an AWS IAM credential for a Google access token with the Security Token Service API?

Yes. POST to /v1/token with grant_type=urn:ietf:params:oauth:grant-type:token-exchange and subject_token_type=urn:ietf:params:aws:token-type:aws4_request, passing a signed AWS GetCallerIdentity request as the subject token. Google validates the signature against the configured AWS workload identity pool provider and returns a federated access token.

What are the rate limits for the Security Token Service API?

Google does not publish a fixed quota for the STS token endpoint, but it is governed by the standard Google Cloud per-project quota system and generally allows several hundred token exchanges per second per project. Tokens are typically valid for one hour, so caching the returned access token until near expiry is the recommended pattern rather than calling /v1/token on every request.

How do I set up workload identity federation through Jentic?

Search Jentic for 'exchange aws credential for google token' or 'oidc token exchange google', load the schema for POST /v1/token, and execute it with your subject_token and audience. Jentic returns the federated access_token which you then pass as a Bearer token to any downstream Google Cloud API call. Get started with Jentic One, the self-hosted execution layer.

Is the Security Token Service API free?

Yes, the STS token-exchange endpoint itself is free to call. You only pay for the downstream Google Cloud services that the returned access token is used to invoke (Cloud Storage, BigQuery, Compute Engine, etc.) at their normal rates.

Does the Security Token Service API support SAML 2.0 assertions?

Yes. Set subject_token_type=urn:ietf:params:oauth:token-type:saml2 and pass a base64-encoded SAML assertion as the subject_token. The workload identity pool provider must be configured with the SAML IdP's metadata before exchange will succeed.

Can I limit what my agent is allowed to do with the Google STS API?

Yes. The Google Security Token Service exposes a single token-exchange operation (POST /v1/token), and because you run Jentic One yourself, your own rules decide which operations and credentials the agent may use. You can restrict the agent to just that token-exchange call so it can mint a federated Google access token from an AWS, OIDC, or SAML subject token and nothing else. The subject credential and audience you configure are injected at execution time and stay outside the agent's context.