intervals-icu-mcp

Overview Schema Source Related Servers Discussions Admin

icu_get_activity_streams

Read-onlyIdempotent

Fetch raw per-sample time-series arrays (power, heart rate, cadence, GPS, etc.) for a single activity. Use for visualization or custom signal analysis when summary metrics are insufficient.

Instructions

Fetch RAW per-sample time-series of one activity — second-by-second arrays for power, HR, cadence, speed, altitude, GPS, temperature, grade, etc.

Heavy payload. Use only when you need the underlying signal for visualization or custom analysis. Most "how was my ride?" questions are better answered by get_activity_details (summary metrics) or get_activity_intervals (per-lap breakdown).

Stream-type filter accepts any of: watts, heartrate, cadence, velocity_smooth, altitude, distance, time, latlng, temp, moving, grade_smooth, raw_heartrate, fixed_heartrate.

latlng comes back as [lat, lng] pairs (null where the sample has no GPS fix).

heartrate is CORRECTED data: Intervals.icu replaces readings above the athlete's configured max HR with an interpolated line at import, so it can never exceed that setting. raw_heartrate is the uncorrected trace and is the only way to see a genuine new max that was clipped away — the platform never raises the setting on its own. Request it explicitly; it is not in the default set.

Input Schema

TableJSON Schema

Name	Required	Description
`streams`	No	List of stream types (e.g., ['watts', 'heartrate', 'cadence']). If not specified, the activity's default streams are fetched — which does NOT include raw_heartrate or fixed_heartrate; ask for those by name.
`max_points`	No	Thin every returned stream to at most this many samples by keeping every Nth one (the same N for all streams, so indices stay aligned). Omit for full resolution. ~500 is plenty to trace a route or see the shape of a ride. Samples are not always 1 s apart, so include 'time' to know when each kept sample was recorded.
`activity_id`	Yes	Activity ID to fetch streams for

Output Schema

TableJSON Schema

Name	Required	Description	Default
`result`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedv5.4.0Sep 30, 2026

addedInput schema / properties / max_points

Added value: +{
+  "anyOf": [
+    {
+      "type": "integer"
+    },
+    {
+      "type": "null"
+    }
+  ],
+  "default": null,
+  "description": "Thin every returned stream to at most this many samples by keeping every Nth one (the same N for all streams, so indices stay aligned). Omit for full resolution. ~500 is plenty to trace a route or see the shape of a ride. Samples are not always 1 s apart, so include 'time' to know when each kept sample was recorded."
+}

Changed1 schema field changedv5.0.1Sep 11, 2026

changedInput schema / properties / streams / description

Previous value: -"List of stream types (e.g., ['watts', 'heartrate', 'cadence']). If not specified, all streams are fetched."New value: +"List of stream types (e.g., ['watts', 'heartrate', 'cadence']). If not specified, the activity's default streams are fetched — which does NOT include raw_heartrate or fixed_heartrate; ask for those by name."

First observedv0.1.0May 19, 2026

TDQS

A4.6/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare this a safe, idempotent, read-only operation, so the safety profile is covered. The description adds genuinely non-obvious behavior beyond them: the payload is flagged as heavy, and it discloses that 'heartrate' is corrected data clipped at the athlete's configured max while 'raw_heartrate' is the only uncorrected trace and the platform never raises the setting. That data-integrity nuance is the kind of trait an agent cannot infer elsewhere.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

It is front-loaded with the purpose, then the payload/usage warning, then the stream list and edge-case notes — a sensible ordering where each block is actionable. It runs long and mildly repeats the schema's own raw_heartrate note, keeping it just short of a 5.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema exists, so return values need not be explained, and the description covers the remaining gaps for a complex tool: payload weight, stream-type vocabulary, default-set behavior, and the corrected-vs-raw HR distinction. An agent has everything needed to call it correctly and interpret the result.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3, and the schema already documents the streams default and the raw_heartrate caveat. The description adds value beyond it by enumerating the full set of accepted stream names (the schema has no enum) and defining the latlng [lat, lng]-with-null format, which raises it above baseline.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource — 'Fetch RAW per-sample time-series of one activity' — and immediately enumerates the concrete arrays returned (power, HR, cadence, speed, altitude, GPS, temperature, grade). It explicitly distinguishes itself from get_activity_details and get_activity_intervals by contrast, so an agent can place it without opening a schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives both when-to-use ('only when you need the underlying signal for visualization or custom analysis') and when-not, naming the two alternative tools and the question types they serve ('how was my ride?'). Nothing is left to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server