[AI] Agentic coaching — where endurance training is heading, and why Tempo was built for it

What’s new

There is a new post on the Tempo blog: Agentic coaching: where endurance training is heading, and why Tempo was built for it. It puts a name to what many of you are already doing with the MCP connector: an assistant like Claude or ChatGPT that reads your rides, wellness and fueling data, follows Tempo’s plan-building protocols, puts the workouts on your calendar, and checks back the following week. The post explains why a chat-only plan falls short, why we built Tempo as the platform under your agent rather than as another chatbot, and the three design choices behind the connector: read before you write, one model of the athlete, and the athlete confirms.

Why it helps

If you have wondered whether to trust an AI assistant with your training, or how Tempo’s approach differs from an in-app AI coach or a read-only connector, the post lays out the reasoning in one place, with sources. It also makes the case for what an agent gives a self-coached athlete that a coach usually cannot: it is there at 6 am and 9 pm, it costs a few dollars a month, and every decision comes with its reasoning attached in your own numbers.

How to try it

Read the post at jointempo.ai/blog/agentic-coaching. To try agentic coaching yourself, go to SettingsAI Assistants, select the Claude or ChatGPT tab, and add Tempo as a custom connector. Try asking:

  • “Read my last six weeks and tell me whether I have headroom under my aerobic ceiling or whether I am ceiling-limited, then propose the block that follows from that.”
  • “Build me an eight-week block for the gravel race on my calendar. Work around the vacation note in week four, simulate the load first, and only create the workouts once I confirm.”
  • “Check in on last week: what did I complete, how did my HRV and resting heart rate track against baseline, and what should change this week?”

Note: MCP access requires a Premium subscription.

Example

Ask: “Why sweet spot this block and not VO2max?”

The assistant answers from your own power-duration curve rather than a textbook, and you can keep asking why until the reasoning is yours.