API coverage
MeiliOps aims to cover 100% of Meilisearch's stable API for the latest stable release. Coverage is measured, not estimated: npm run coverage scans the source for every API call and compares it with the operations in Meilisearch's published OpenAPI spec.
Meilisearch 1.54.1: 143 operations. 140 have a dedicated screen (97.9%); all 143 are reachable from the API console.
| Area (OpenAPI tag) | Operations | Where in MeiliOps |
|---|---|---|
| Settings | 69 | Index Settings tab (every sub-route, plus All settings) |
| Async task management | 10 | Tasks, Batches, Activity |
| Documents | 10 | Documents tab, Edit with function |
| Indexes | 9 | Sidebar, Index info tab (create, rename, primary key, swap, compact, delete) |
| Chats | 7 | Chats |
| Search rules | 5 | Search rules |
| Experimental features | 5 | Experimental; the 3 /network routes are console-only (see below) |
| Keys | 5 | API keys |
| Webhooks | 5 | Webhooks |
| Logs | 3 | Logs |
| Backups | 2 | Overview (dumps and snapshots) |
| Search, Similar documents, Facet search | 5 | Search tab |
| Stats | 2 | Overview, Index info, Metrics |
| Multi-search | 1 | Multi-search |
| Export | 1 | Export |
| Template | 1 | Test template… in Settings |
| MCP connection | 1 | Overview MCP card |
| Health, Version | 2 | Connection test, Overview |
Not covered by a dedicated screen
GET /network, PATCH /network and POST /network/control configure sharding across a cluster, which is an Enterprise Edition feature. They're out of scope for a dedicated screen and available from the API console.
Experimental routes
Experimental routes are covered on a best-effort basis. Some need a launch flag on the server, others a runtime toggle; MeiliOps tells you which when a route isn't enabled.
Keeping it at 100%
CI runs npm run coverage -- --strict on every push. It fails when a new stable operation appears in the spec without a screen, so a new Meilisearch release can't silently reduce coverage. See New Meilisearch versions.