{"server":{"$schema":"https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json","name":"com.scorecompute/scorecompute","description":"27 MCP tools for independent Rust computation, Discovery studies and typed native composition.","title":"ScoreCompute","version":"0.7.0","websiteUrl":"https://scorecompute.com/","icons":[{"src":"https://scorecompute.com/favicon.svg","mimeType":"image/svg+xml"}],"remotes":[{"type":"streamable-http","url":"https://scorecompute.com/mcp"}],"_meta":{"io.modelcontextprotocol.registry/publisher-provided":{"authentication":"No API key required during the rate-limited public alpha.","availability":"Public alpha backed by a locally operated compute machine. Calculations can be temporarily unavailable when that machine or its connection is offline. The separate contributor pilot depends on opted-in clients and exposes queued/expired states.","categories":["Scientific computing","Statistics","Optimization","Astronomy","Physical consistency"],"gpuTools":["simulate_pi"],"limits":{"maxConcurrentRustCalculations":1,"maxRequestBytes":1048576,"maxSessionsPerAddress":8,"maxSessionsTotal":64,"sessionIdleMinutes":15},"privacy":"The public observatory shows declared MCP client name/version, tool name, status and execution duration. It does not publish scientific inputs, observations, candidate IDs, claims, perturbations, intermediate fits or results. Do not put personal information or secrets in clientInfo. A measured/synthetic data-origin declaration is supplied by the caller and is not independently verified. Contributor jobs expose synthetic workload/scenario IDs, seed/sample parameters and receipts by opaque ticket; pool status shows chosen aliases, hardware offers and supported workloads, never worker tokens or raw hardware identifiers.","summary":"27 MCP tools: fourteen base Rust computation tools, four independent Discovery study tools, six optional capability and mission tools, and three contributor-network tools. The typed planner exposes 21 implemented capabilities and 44 versioned fact contracts; the narrower legacy list_capabilities catalogue retains 13 engine contracts. plan_observation derives a fit from supplied angular observations and predicts conditional uncertainty for declared candidate series; a forecast is not acquired data. stress_test_observation tests a finite declared list of hypothetical perturbations; absence of a counterexample is not global robustness. assimilate_observation refits original and supplied new angles once, keeping caller-declared synthetic/measured origin unverified. improve_schedule compares permitted changes and checks one finalist with a separately seeded confirmation. All four can be called without the composer. The planner can bind complete Discovery requests but cannot turn predicted angles into acquired observations. Approved task and route risk chains retain checked duration adapters, explicit units and native events. Structural graph cost is not runtime, scientific quality or price. CUDA applies to simulate_pi and the distinct public pi_chunk_v1 contributor workload. Client 0.2.0 retains sc-worker/0.1 negotiation and adds only three compiled schedule_probe_v1 CPU scenarios. Workers declare supported workloads; throughput is reported separately by workload. All contributed counts are recomputed on coordinator CPU, so no net speedup is claimed. Scientific inputs/results stay private to their requester; public contributor tickets contain only synthetic parameters and receipts. Public alpha; conditional findings are never authenticity or misconduct verdicts.","tags":["rust","cuda","monte-carlo","optimization","astronomy","solar-shadows","statistical-audit","mcp","native-orchestration","robust-scheduling","typed-composition","capability-planning","geometric-routing","route-risk","distributed-computing","volunteer-computing","observation-design","conditional-uncertainty","counterexample-search","data-assimilation","schedule-improvement"],"toolNames":["simulate_pi","optimize_tasks","compute_orbit","verify_shadows","check_ephemeris","audit_statistics","analyze_kinematics","check_shadow_track","solve_equilibrium","analyze_chess_game","screen_wash_trading","simulate_plan","route_optimizer","triangulate_tracks","plan_observation","stress_test_observation","assimilate_observation","improve_schedule","list_capabilities","plan_mission","build_robust_plan","list_composition_capabilities","compose_capabilities","execute_composition","network_status","submit_network_job","get_network_job"]}}},"_meta":{"io.modelcontextprotocol.registry/official":{"status":"active","statusChangedAt":"2026-09-30T21:08:24.619632Z","publishedAt":"2026-09-30T21:08:24.619632Z","updatedAt":"2026-09-30T21:08:24.619632Z","isLatest":true}}}