Klarna is Poor to agents.
Discry independently scored how well an AI agent can discover and understand the Klarna 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 · 40/100Comprehension
55% of score · 63/100What we found
- Klarna has comprehensive multi-step workflow documentation showing the full payment integration flow from initiation through order management
- An OpenAPI spec exists as a GitHub gist but is not officially hosted on docs.klarna.com, making agent discovery unreliable
- A Klarna Payments MCP server exists on Glama, auto-generated via AG2 MCP builder
- Documentation includes dedicated error handling pages with error codes per product (Payments, Order Management)
- The API reference landing page returns 404, suggesting documentation restructuring may be in progress
What to change
Prioritized by impact on discoverability. You (or your docs platform) deploy these — Discry never touches your API.
- 01Add llms.txt to docs.klarna.com covering core payment flows and API endpoints
- 02Host the OpenAPI spec officially at a public URL on docs.klarna.com
- 03Add AGENTS.md to the klarna GitHub org with guidance for payment integration agents
- 04Fix the API reference introduction page (currently 404)
- 05Add .well-known/mcp.json pointing to the Klarna MCP server on Glama
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.
Klarna uses API key auth and OAuth2. JSON error responses with error codes documented per product. Rate limits documented. Pagination supported.