Ghost scores 77.1 on the Kin Score, exemplar band, down 1.3 from 78.4. Its Agent Readiness is 30.8, agent-ready. That is a strong record for an MIT-licensed publishing platform that funds a non-profit foundation, and most of it is earned by Ghost itself. The part that is not clean is ours: the catalog shows 33 API pages for a provider that publishes 3 contracts.
What is in the surface
Every Ghost site exposes two documented REST APIs under https://{site}/ghost/api/. The Content API is read-only and key-authenticated, for delivering posts, pages, tags, authors, tiers and settings to front-ends. The Admin API is read-write and JWT-authenticated, for managing posts, pages, members, tags, tiers, offers, newsletters, users, media, themes and webhooks.
The provider record says source_spec_count: 3 and split_spec_count: 33. Split by tag, those three files became six Content pages holding 14 operations, thirteen admin- prefixed pages holding 42, and fourteen unprefixed pages that cover the same Admin resources again with 48. The agentic-access artifact counts 56 operations, 26 acting, which is exactly 42 plus 14. So the 56 is the real surface, and the fourteen unprefixed pages are a second copy of the Admin API from an earlier split that was never retired. A reader browsing the provider page sees Tags API (6 operations) next to Ghost Admin - Tags API (4) and has no way to tell which one is current. That defect is ours, not Ghost’s.
| Facet | Score |
|---|---|
| Operational transparency | 89.5 |
| Developer ergonomics | 86.9 |
| Discoverability | 80.0 |
| Access clarity | 76.3 |
| Contract quality | 66.7 |
| Contract governance | 31.8 |
Transparency and ergonomics carry the score, and the open-source facet scores 100. Contract governance is the weak column, and the upsert check scores 0. All 33 contracts score as callable even though they are templated on the customer’s own domain, which is honest: every Ghost install is its own host.
The agent story
Ghost ships first-party MCP. A hosted, anonymous, read-only documentation server runs at docs.ghost.org/mcp and answers a live tools/list, and a separate local server built into ghst, Ghost’s CLI, exposes the Admin API as roughly 90 tools once an operator has authenticated it. The catalog also probed an A2A agent card at docs.ghost.org/.well-known/agent-card.json on 2026-09-26 and graded it conformant at protocol 0.3.
Yet the Agent Readiness record has agent_card: false. Either the dimension is measuring a different host than the probe found, or the probe result never reached the scorer. In both cases the provider published the artifact and the score does not show it. I am reporting that as a catalog defect rather than a Ghost gap. The other dark dimensions are real: no idempotency, no protected resource metadata, no dynamic client registration, no reversibility notes, and no examples in the contracts. Agentic access is derived, so I do not credit Ghost for it.
Takeaway
On our side: retire the fourteen duplicate Admin pages and reconcile the agent card. On Ghost’s side, three things would move the number. It could publish its own OpenAPI for both APIs so the contract facets stop resting on our splits. It could add request and response examples. And it could document idempotency on the Admin writes, which matters for an agent publishing posts or changing members. The documentation and the open-source work are already exemplar. The contract layer is what is left.
See the full profile at apis.io/providers/ghost-org/.