Braintree is Poor to agents.
Discry independently scored how well an AI agent can discover and understand the Braintree API from what’s public — not whether it’s usable. Below: every signal we checked, what’s costing the score, and what to change.
SCORED UNDER RUBRIC 1.2 · A full re-launch under Discry Score 2.5 — a new behavioral instrument, not comparable to these scores — is in progress.
Discovery
45% of score · 43/100Comprehension
55% of score · 69/100What we found
- Braintree has a dedicated llms.txt (at developer.paypal.com/braintree/llms.txt) with structured sections and the same markdown URL hint as PayPal (/md/ prefix), making docs agent-accessible.
- As a PayPal subsidiary, Braintree inherits the parent's robots.txt but has no independent discovery signals — no own domain, no AGENTS.md, no MCP server.
- Braintree uses GraphQL as its primary API (not REST), with the schema publicly available on GitHub, but this format is less standard for agent tooling than OpenAPI.
- No dedicated Braintree MCP server exists in any registry — agents must use PayPal's broader toolkit or integrate manually.
- Tutorial-focused documentation with step-by-step guides (Drop-in UI, Hosted Fields) provides good workflow coverage for agents learning to integrate.
What to change
Prioritized by impact on discoverability. You (or your docs platform) deploy these — Discry never touches your API.
- 01Publish a Braintree-specific MCP server or ensure the PayPal Agent Toolkit explicitly covers Braintree GraphQL operations.
- 02Create an OpenAPI-equivalent description of the GraphQL schema that agents can more easily parse (or a GraphQL-specific llms.txt section).
- 03Add AGENTS.md to the braintree/graphql-api GitHub repo to guide coding agents.
- 04Document rate limits publicly — agents have no way to know when they'll hit throttling.
- 05Add idempotency documentation for the GraphQL API to help agents avoid duplicate operations.
Execution coverage · INFORMATIONAL, UNSCORED
Whether an agent can actually complete a call and recover from errors is the deeper Audit layer — documented here, but not part of the Discry Score.
Authentication via public/private key pair for server-side and client tokens for frontend. GraphQL API uses structured error responses. Rate limits not publicly documented. Pagination via cursor in GraphQL. No idempotency key documentation found.