MCP Server for FFBB (Fédération Française de Basketball) data — scores, standings, calendars and live data for AI agents.
The FFBB MCP server provides 2 tools with mixed quality. Naming is action-oriented (ffbb_search, ffbb_version) and follows verb_noun convention. Both tools have substantial descriptions that explain purpose and use cases. However, the schema documentation is incomplete, while ffbb_search has enumerated type constraints and detailed parameter descriptions, the input parameter objects lack explicit JSON Schema type declarations visible in the source. ffbb_version declares an empty input object {} with no parameters, which is correct but minimal. The ffbb_search tool demonstrates good constraint design with enum values ('all', 'competitions', 'organismes', etc.) and descriptive parameter guidance. Error handling and output schemas are not visible in the provided code excerpts, which limits assessment of recovery guidance and structured output practices. The server includes tool annotations (readOnlyHint) for both tools, indicating awareness of current MCP patterns, but overall schema rigor could be stronger.
Recherche FFBB (multi-index) — texte libre sur compétitions, organismes, rencontres, salles, pratiques, terrains, tournois, engagements, formations, officiels, entraîneurs, communes.
Informations de version et configuration runtime du serveur FFBB MCP. Retourne `dict` compact et typé `{package_version, mcp_sdk_version, python_version, transport, cache_ttls}` ; lecture seule, idempotent, sans appel réseau externe ni effet de bord, <10ms. Utilise cet outil pour diagnostiquer la version déployée, vérifier le transport (`stdio` vs `streamable-http`) ou inspecter les TTL avant de debugger un autre outil. Ne pas utiliser pour obtenir des données basket — utilise `ffbb_search`, `ffbb_club` ou `ffbb_bilan` à la place ; pour la santé live, préfère `/health`.
Output schemas not documented in provided source excerpts. LLMs cannot plan downstream tool calls or extract specific fields without knowing what ffbb_search and ffbb_version return.
ffbb_search parameter 'limit' lacks min/max constraints. Unbounded integer can allow LLMs to pass absurd values (0, 999999) that break pagination or timeout.
ffbb_search 'filter_by' parameter lacks specificity. Description says 'Filtre optionnel Meilisearch' but does not explain Meilisearch filter syntax, valid operators, or examples. LLMs cannot use this correctly.
Error handling guidance not visible in provided code. No evidence of recovery patterns (pattern:recovery-guide), error classification, or actionable error messages to guide LLM retry logic.
Inferred effective spec: 2026-07-28+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | B | 73 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 34 | - | v1 |
Tool composition risk: ffbb_search returns results but downstream tool availability unclear. If agents must call additional tools (e.g., ffbb_club, ffbb_bilan mentioned in ffbb_version description) to act on search results, output should include IDs/references those tools accept.