Microsoft Azure InstanceMetadataClient

Name: InstanceMetadataClient API
Brand: InstanceMetadataClient
Availability: InStock

★ Only Publicly Available OpenAPI DocumentCloud InfrastructureComputebasic4 Endpoints

Install Jentic One Beta

Connect the InstanceMetadataClient to your agent

Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the InstanceMetadataClient, 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 InstanceMetadataClient API.

Retrieve compute, network, and tag metadata for the running Azure VM

Fetch an attested document signed by Azure to prove VM identity to relying parties

Look up the managed identity client and tenant ids assigned to the VM

GET STARTED

Start building with InstanceMetadataClient API

Explore with Jentic

View OpenAPI Document

For Agents

Read Azure VM instance metadata and obtain managed identity OAuth tokens from inside the VM via the Azure Instance Metadata Service.

Use for: Get the metadata of the Azure VM I am running on, Retrieve a managed identity access token for an Azure resource, Find the client id of the managed identity assigned to this VM, Get an attested document to prove this VM's identity

Not supported: Does not handle Azure Resource Manager calls, secret storage, or cross-VM discovery - use for in-VM metadata reads and managed identity token issuance only.

Use Cases

Patterns agents use InstanceMetadataClient API for, with concrete tasks.

★ Credential-Free Azure SDK Auth

Workloads running on Azure VMs avoid embedded secrets by calling /identity/oauth2/token to obtain a managed identity bearer token, then passing it to other Azure APIs. The IMDS endpoint is reachable only from the VM itself, so no inbound network rules are needed. Adopting this pattern typically takes a couple of hours per workload and removes a class of secret-rotation problems.

Call /identity/oauth2/token with resource=https://management.azure.com/ from the VM and return the bearer token.

VM Self-Description for Tagging and Routing

Bootstrap scripts read /instance to discover the VM's region, vmSize, resourceGroupName, and tags, then configure their behaviour accordingly - for example, registering with a region-specific service discovery system. Because IMDS is local, the call is fast and does not depend on AAD. This is the standard way to make Azure VMs self-aware without baking config into images.

GET /instance and return the compute.location, compute.vmSize, and the value of the tag named 'environment'.

Attested VM Identity for Zero-Trust

Relying parties verify VM identity by fetching the /attested/document, which returns a JWT signed by Azure that includes the subscriptionId, vmId, and other compute fields. The receiving service validates the signature and checks the claims before granting access. This is the foundation for zero-trust admission to internal services that should only accept calls from sanctioned Azure VMs.

GET /attested/document?api-version=2019-08-15 and return the signed JWT.

Agent-Driven IMDS Calls via Jentic

An AI agent running on (or proxying to) an Azure VM can fetch metadata and tokens through Jentic without code that hard-codes the IMDS host. The agent searches Jentic for 'get an azure managed identity token', loads the operation schema, and executes against the local IMDS. Jentic mediates the request shape and surfaces the response into the agent's context.

Use the Jentic MCP tool azure_imds_get_token to fetch a managed identity token for resource https://vault.azure.net and return the access_token.

Key Endpoints

4 endpoints — jentic publishes the only available openapi specification for instancemetadataclient, keeping it validated and agent-ready.

METHOD

PATH

DESCRIPTION

GET

/instance

Return compute, network, and tag metadata for the running VM

GET

/identity/info

Return managed identity client and tenant ids for the VM

GET

/identity/oauth2/token

Issue a managed identity OAuth 2.0 access token for a target resource

GET

/attested/document

Return a signed attested document for the VM

GET

/instance

Return compute, network, and tag metadata for the running VM

GET

/identity/info

Return managed identity client and tenant ids for the VM

GET

/identity/oauth2/token

Issue a managed identity OAuth 2.0 access token for a target resource

GET

/attested/document

Return a signed attested document for the VM

Why Jentic?

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

Credential isolation

IMDS itself needs no stored credential because it trusts the in-VM origin, and any managed-identity token it issues is held 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 'get an Azure managed identity token' or 'read this VM's instance metadata', and Jentic returns the matching IMDS operation with its input schema so the agent calls the right endpoint without remembering the link-local URL or required headers.

Time to first call

FAQs

Specific to using InstanceMetadataClient API through Jentic.

Why is there no official OpenAPI spec for InstanceMetadataClient?

Microsoft documents IMDS as a REST surface but does not publish a maintained OpenAPI 3 specification for it. Jentic generates and maintains this spec so that AI agents and developers can call InstanceMetadataClient via structured tooling. It is validated against the live API and kept up to date. Get started with Jentic One, the self-hosted execution layer.

What authentication does the InstanceMetadataClient use?

IMDS itself is reachable only from the VM at 169.254.169.254 and requires the request header Metadata: true on every call. Through Jentic the request shape including this header is handled automatically. The /identity/oauth2/token endpoint then issues a bearer token for downstream Azure services.

Can I get a managed identity token from this API?

Yes. GET /identity/oauth2/token with the api-version query parameter and a resource parameter (e.g. resource=https://management.azure.com/) returns an access_token, expires_in, and the client_id of the identity. Use the token as a Bearer credential for the target Azure service.

What are the rate limits for IMDS?

IMDS is a per-VM local service and does not publish a per-second rate cap, but Microsoft recommends caching tokens until close to expiry rather than calling /identity/oauth2/token on every request. Treat repeated calls within a few seconds as wasteful, not as a hard limit.

How do I retrieve VM metadata through Jentic?

Run pip install jentic, then search Jentic with 'get azure vm instance metadata', load the /instance operation, and execute it. Jentic returns the compute, network, and tag JSON exactly as IMDS produces it, so downstream code reads .compute.location and friends directly.

Does IMDS work from outside the VM?

No. The endpoint 169.254.169.254 is link-local and only reachable from inside the Azure VM. For workloads outside the VM that need similar token issuance, use Azure AD app registrations or workload identity federation instead.

Can I limit what my agent is allowed to do with the Azure Instance Metadata Service?

Yes. Because you run Jentic One yourself, your own rules decide which IMDS operations the agent may call, so you can allow only reading instance metadata from /instance while blocking the managed-identity token request at /identity/oauth2/token, or the reverse. The attested document read at /attested/document is not available to the agent unless you explicitly add it. IMDS trusts the in-VM origin rather than a stored credential, and any managed-identity token it issues is held by your own Jentic One instance and injected at execution time rather than exposed to the agent.