Skip to content

Talk the plan through on your laptop. Train it on your phone.

Your AI agent plans programmes, workouts and exercises with you, then saves them straight into StrongLog. Nothing to copy, nothing to retype.

you I want three short strength days a week, but some weeks I only make it twice.

stronglog_training_context · no active programme

agent Then a flexible A/B/C sequence fits: squat and hinge, press and pull, deadlift and overhead. A light week just means the next day waits for you.

stronglog_programme_evaluate · 3 sessions · 0 issues

you Looks right. Save it.

stronglog_programme_apply · saved · receipt 7f3a…

StrongLog Today screen showing the Squat and hinge workout ready to start
Example data · fictional athlete
  1. Discuss
  2. Evaluate
  3. Save
  4. Train
  5. Log

From conversation to barbell in five steps.

StrongLog connects to agents over MCP, the open standard for giving an AI assistant tools. Your agent plans with you; your phone runs the session. Screens below show example data from a fictional athlete.

  1. 01 · Discuss

    Start with how you actually train.

    Tell your agent what you want and what your weeks really look like. It reads your current programmes, progression and recent sessions from StrongLog before suggesting anything.

    you Three strength days a week, but some weeks I only make it twice.

    stronglog_training_context · no active programme

    Nothing saved yet.

    Your agent is still drafting and checking. StrongLog changes only when you ask it to save.

  2. 02 · Evaluate

    The draft is checked before anything changes.

    The agent runs the programme through the same validator the app uses. Problems come back as issues to fix in the conversation. Your training stays untouched.

    agent A flexible A/B/C sequence fits. Each day comes up when you train, not when the calendar moves.

    stronglog_programme_evaluate · 3 sessions · 0 issues

    Nothing saved yet.

    Your agent is still drafting and checking. StrongLog changes only when you ask it to save.

  3. 03 · Save

    Say save. That is the approval.

    You allow your agent to plan once, when you connect it. After that, there is no separate approval visit for each workout. Every save returns a receipt you can look up later.

    you Looks right. Save it.

    stronglog_programme_apply · saved · receipt 7f3a…

    StrongLog Programmes screen with the saved Autumn Base programme marked primary
  4. 04 · Train

    Pick up your phone. The next workout is there.

    No copying, no retyping. The programme you discussed is already in your plan with its exercises, sets and starting weights.

    StrongLog Today screen showing the Squat and hinge workout ready to start
  5. 05 · Log

    Log the sets. The history stays yours.

    StrongLog works offline in the gym. Completed workouts are records of what happened, and your agent reads them back next time you plan.

    A live StrongLog workout with sets checked off and a rest timer running
StrongLog recommending recovery without skipping or moving planned training
Example data · fictional athlete

Built for the weeks that do not go to plan.

Agents are good at planning. StrongLog makes sure the plan survives real life, without quietly rewriting what you did.

Your schedule, your rules

Run a sequence whenever you can train, aim for a number of days a week, or fix sessions to set days. Change the schedule later without rebuilding the programme.

Time passing is not failure

A flexible programme only moves forward when you train or decide to move it. A quiet week does not pile up missed workouts.

History stays history

Completed sessions are facts. Recommendations explain themselves, and corrections leave a clear trail instead of overwriting the record.

Works in the gym

Install it on your phone and log without signal. It catches up when you are back online.

Bring the agent you already use.

Any assistant that supports remote MCP servers with OAuth sign-in can connect. You choose what it may do: read your training, plan future sessions, or correct logged history.

server    https://your-stronglog-host/api/agent/v1/mcp
transport Streamable HTTP
sign-in   OAuth with PKCE
scopes    training.read
          training.plan.write
          training.log.write