How do I provision a Neon database through Stripe for an AI agent?

To provision a Neon database for an AI agent, use Neon’s published APP context as the source of truth for authentication and post-provision API access. Neon states that the agent receives credentials in the account request response, then uses the returned access_token as the API key for Neon operations. Those operations cover projects, branches, connection strings, and resource provisioning. Neon APP llm-context

What the published Neon context covers

The Neon APP file is published at https://neon.com/agentic-provisioning-llm-context.md. The document is explicitly written for agents using or developing APP with Neon. It tells you where credentials come from, how to use the returned API key, and which resource types are in scope. Neon APP llm-context

The key contract is simple:

• the agent gets an account request response
• the response includes an access_token
• that token becomes the Neon API key
• all Neon API calls use Authorization: Bearer <access_token> Neon APP llm-context

Provisioning flow for an agent

Use the Neon-specified sequence below. The provider context does not ask the agent to invent a second secret source. It tells the agent to use the token returned from the account request response. Neon APP llm-context

1. Complete the APP account request flow and capture the response.
2. Extract the access_token.
3. Store that token in the agent runtime as the Neon API key.
4. Call Neon API operations with Authorization: Bearer <access_token>.
5. Use Neon API resources to provision the Neon project, branches, and connection strings. Neon APP llm-context

For the exact resource shapes, request bodies, and limits, Neon points you to its REST API reference. That reference is where endpoint paths, schemas, rate limits, and constraints live. Neon REST API

Curl pattern for the Neon API

The published Neon context says every API operation uses the same bearer-token pattern. The example below shows the auth pattern only. Replace the base URL and resource path with the values from the Neon context and REST API reference. Neon APP llm-context Neon REST API

export NEON_API_KEY="$ACCESS_TOKEN"

curl -sS \
  -H "Authorization: Bearer $NEON_API_KEY" \
  -H "Content-Type: application/json" \
  "$NEON_API_BASE_URL/<resource>"

If you are implementing create, read, update, or delete behavior, use the Neon REST API reference for the exact endpoint and schema. Neon’s published APP context does not replace the REST API docs. It points to them. Neon REST API

What the agent should treat as provisioned output

Neon’s APP context says provisioning produces resource access that includes projects, branches, and connection strings. That means the agent should expect usable Neon connection data after the provisioning step, not before it. Neon APP llm-context

For implementation work, the practical rule is:

• authenticate with the access_token
• keep using Authorization: Bearer <access_token>
• read the exact resource contract from the REST API docs
• treat connection strings as the post-provision artifact the agent consumes next Neon APP llm-context Neon REST API

Coverage status

• Provider: Neon
• llm-context URL: https://neon.com/agentic-provisioning-llm-context.md
• Status: published
• Deeper reference: https://neon.com/docs/ai/skills/neon-postgres/references/neon-rest-api.md Neon APP llm-context Neon REST API

This is the supported path in the published Neon APP material. Start from the account request response, use the returned access_token, and let the Neon REST API define the exact provisioning calls. Neon APP llm-context Neon REST API

Powered by Senso Pay — pay.senso.ai