icuvisor

Ricardo Niederberger Cabral·RicardoCabral.icuvisor

MCP server connecting intervals.icu training data to AI assistants.

icuvisor is a local-first Model Context Protocol server that connects intervals.icu training data to AI assistants such as Claude, ChatGPT, Pi, Cursor, and other MCP-compatible clients.

winget install --id RicardoCabral.icuvisor --exact --source winget

Latest 1.6.3·August 25, 2026

Release Notes

Added

• Structured workout steps now accept press_lap:true, which serializes the documented Intervals.icu Press lap marker while retaining a required duration or distance for planned load estimates. The workout-syntax resource documents Garmin-focused device support and fidelity warnings remain explicit when the returned workout document omits the control. • get_extended_metrics now returns eight running-dynamics metrics at activity and interval scope when present:stance time, vertical oscillation and ratio, step length, stance-time percent and balance, vertical speed, and leg-spring stiffness. Responses preserve explicit units and source provenance, omit absent/null/malformed values, and retain zero values. average_impact_loading_rate remains raw-only because its unit is not documented.

Changed

• Disabled automatic Winget manifest submission until the fork token can execute UpdateRef; the manual recovery workflow remains available.

Fixed

• Regenerated the tool-input schema snapshots after adding structured-workout press_lap support, restoring the schema freshness CI guard.

Installer type: wix

x64DFEC398AF3714B1DA886BF2797E433A444E1276E173D0194F87BDF64B7AC804E
arm6427B6E5D8165170B374DEDCB6D9F8CB6361D970F4C0C7370D903310F301E8598D

Details

Homepage
https://icuvisor.app/
License
MIT
Publisher
Ricardo Niederberger Cabral
Support
https://github.com/ricardocabral/icuvisor/issues
Moniker
icuvisor

Tags

aicoaching-toolfitnesshealthintervals-icullmllm-integrationmacosmcpmcp-serverwellnesswindows

Older versions (11)

1.5.10
x64BF286BC40FD327A86606781F060EAC53D96BE07114D70BA7CDDC5216398F0B0E
arm6477EF58AA79962A881B239B5B655769A87312891A26BDE7F2AC3B289AE3D51DF3
1.5.4
x64B4314FFEBD46FDCFD1CCEC8D2A27F562950256384376AE9E3245CC6BB3F52DF1
arm6455AB659DFDC2FF7B0F79F0A4C636D213B66F7D9A1391408D7EE0C8F7814B2F91
1.5.3
x642B223CB6537FA790F5D0AC146D640961DCC9A0FE76178E8F49472CA1A41460FE
arm6412ED46E2F2D76F7A3474AFA00490DDAF0405DC511DD70C225124EDFE420F2E13
1.5.2
x64B89B3E73065793B7BE7BE8437F73E48FB474C54BD45AA23F06858C66C05AA1E6
arm644669B3F0CFECF8A44E3EB5F1EDCE4AA363B3FAD0B3DF4B0DFED957175D4D8EEE
1.5.1
x64D9E9F17B13CBECDB08DB1FDABB2FA72CDEA6FCA0788B42B7A221A12AFC7B234B
arm649084F8B682BF95B71D921E5ABBF6E9E7C92CB99882F7019B8F39CF299AC29160
1.5.0
x64FB716CF240E7F4F6C39B067239D8D5CE0D5C3FD1A2A4C3195744EBE2ED85FE71
arm64FD335B12062F52E9FFC116BDE1637905BC23F334CF121BA97885903E41DB2C8A
1.4.0
x64CA6AA4EBE6034AA33D172855FDBEE5A3128093488AF267764F1375DF5CC00CDC
arm6419D27261FAEFCC911B4BB82849D3E706AAE5CDD7983AA07AE89499DAF7552CA6
1.3.0
x64E6DED133C48F108C385DC8879CEC92F724C02E45F9D1F34C4A07F9CA19C7BB9E
arm643BA40BD48D3A5A0BB4A2647E2C898F22ACBB4E487A434058F4C4FDB18163C0E9
1.2.1
x64E7A8EDBC678F917A71C112B4EBB8A9F12C653077E45751B32EEDD26B066D084C
arm64C73E85C683D44F5EE9DA7F8DEF6465A0784802A3DE7A96240BDB9F8A19421912
1.2.0
x640499A896B9F5A50A03BAB710058D625FB9FF505FE12114089EA44D5E5B6DAB96
arm6458AA193EF345C8A26B8C0FCE0AC941837677AA4FB45AECD583E3B94735C52AD9
1.0.0
x64EDB7F6B107E53BB7D753E0B5705D7F5DF622101EBA18FA634922BD45696A8A70
arm64812B54331B353CB48BCCB1876F1CF5F638AA04051DB2B3CD4525A71F6CDE5A97