{"name":"LMCache Space","summary":"A hosted planning and MCP diagnostics workspace for teams deploying LMCache-compatible KV cache layers across LLM inference clusters.","endpoint":"https://lmcache.space/mcp","server_card_url":"https://lmcache.space/.well-known/mcp/server-card.json","authentication":{"type":"bearer","paid_access":true,"token_claim_method":"contact_support_until_checkout_is_configured"},"pricing":{"starter":{"id":"starter","name":"Starter","summary":"One cache-readiness review, sizing estimate, and MCP policy export for a single inference service.","monthly":{"displayMonthlyUsd":49,"dueTodayUsd":49,"coverage":"one month","renewsAutomatically":true},"annual":{"displayMonthlyUsd":24.5,"dueTodayUsd":294,"coverage":"one year","discount":"50%","renewsAutomatically":true}},"team":{"id":"team","name":"Team","summary":"Repeated KV cache sizing, rollout evidence, and MCP cache-policy exports for a platform team.","monthly":{"displayMonthlyUsd":149,"dueTodayUsd":149,"coverage":"one month","renewsAutomatically":true},"annual":{"displayMonthlyUsd":74.5,"dueTodayUsd":894,"coverage":"one year","discount":"50%","renewsAutomatically":true}},"scale":{"id":"scale","name":"Scale","summary":"Portfolio cache governance, multi-cluster rollout planning, and trace review for production teams.","monthly":{"displayMonthlyUsd":399,"dueTodayUsd":399,"coverage":"one month","renewsAutomatically":true},"annual":{"displayMonthlyUsd":199.5,"dueTodayUsd":2394,"coverage":"one year","discount":"50%","renewsAutomatically":true}}},"tools":[{"name":"estimate_kv_cache_savings","description":"Estimate reused prompt tokens, prefill time avoided, and GPU spend saved by a KV cache layer.","input_schema":{"type":"object","properties":{"input":{"type":"string"},"mode":{"type":"string"}},"required":["input"]},"output_schema":{"type":"object","properties":{"ok":{"type":"boolean"},"summary":{"type":"array","items":{"type":"string"}},"next_action":{"type":"string"}}}},{"name":"plan_lmcache_rollout","description":"Create a rollout checklist for vLLM, SGLang, storage tiers, observability, and cache invalidation.","input_schema":{"type":"object","properties":{"input":{"type":"string"},"mode":{"type":"string"}},"required":["input"]},"output_schema":{"type":"object","properties":{"ok":{"type":"boolean"},"summary":{"type":"array","items":{"type":"string"}},"next_action":{"type":"string"}}}},{"name":"inspect_inference_trace","description":"Review trace summaries for cache-hit rate, long-context pressure, and risky miss patterns.","input_schema":{"type":"object","properties":{"input":{"type":"string"},"mode":{"type":"string"}},"required":["input"]},"output_schema":{"type":"object","properties":{"ok":{"type":"boolean"},"summary":{"type":"array","items":{"type":"string"}},"next_action":{"type":"string"}}}},{"name":"generate_mcp_cache_policy","description":"Produce an agent-readable cache policy for retrieval, routing, and long-context inference workflows.","input_schema":{"type":"object","properties":{"input":{"type":"string"},"mode":{"type":"string"}},"required":["input"]},"output_schema":{"type":"object","properties":{"ok":{"type":"boolean"},"summary":{"type":"array","items":{"type":"string"}},"next_action":{"type":"string"}}}}],"client_guides":["/mcp/claude-desktop/","/mcp/cursor/","/mcp/vscode/","/mcp/chatgpt/"]}