smspva-docs
Score by pillar
reliability: LOW (T1 single-shot, not credited - requires T3 >=24h)
Context-cost
tools/listPaid on every request. Lean footprint for 4 tools.
MCP primitive coverage
the quality score covers toolsWhy this score — causal attribution
Each penalty: measure → mechanism → effect → Δscore.
[MCP06] lethal trifecta: untrusted-content ingestion + sensitive-data access + exfil/destruction -> an injection can exfiltrate → MCP security flaw → tool: (serveur)
50% typed+described params → schema = the contract the agent fills in → more malformed calls
proto 2024-11-05 (4 rev. behind 2025-11-25) → outdated protocol version → recent features unavailable, recent clients weakened
no OAuth discovery (.well-known missing) → RFC 9728/8414 not exposed → auth not auto-discoverable by clients
Security — OWASP MCP Top 10
Combines untrusted-content ingestion (1), sensitive-data access (2) and exfiltration/destruction (4). A prompt-injection could read a secret then exfiltrate it.
Fix: scinder en 2 serveurs : un MCP « lecture/contenu non-fiable » (read-only, openWorldHint) isolé du MCP « privilégié » (données sensibles + exfil/destruction).
(serveur)Badge — paste it in your README
[](https://checkmcp.dev/report?url=https%3A%2F%2Fmcp.smspva.com%2Fdocs)