The questions a spreadsheet can't answer.
Your assistant reads your real training — sessions, trends, recovery, plans — and answers in words, not formulas.
“Am I actually getting stronger on squat this block?”
“What should my next session look like given how I've recovered?”
“Compare my bench with and without Kaatsu.”
“Where am I losing bar speed — and on which lifts?”
“Draft a deload week and show me the changes before I accept.”
“How did my sleep line up with my heavy days this month?”
An assistant that knows your training.
Because it's connected to your actual data over MCP, it does more than chat — it reads, analyzes, and proposes.
It reads your real numbers.
Not generic advice — your sessions, your per-rep velocity and power, your estimated maxes, your recovery. Answers grounded in what you actually did.
It can draft changes.
Ask for a deload or a swap and your assistant can propose the plan changes for you — laid out for you to accept, edit, or decline. Nothing changes without your say-so.
Honest by default.
The same rule as the rest of Eccentric: sample sizes are shown and patterns are framed as questions, never as proof. Your assistant won't invent causation.
Your data, your terms.
The connection is yours to grant and revoke. Your assistant reaches your training only with your permission — and you can take your data out as a CSV anytime.
Plug it in and just ask.
MCP — the Model Context Protocol — is the standard that lets an AI assistant securely reach a tool like Eccentric. You grant the connection; you can pull it anytime.
Connect your assistant
Add Eccentric to an MCP-capable assistant like Claude or ChatGPT. It's a one-time connection you control.
Ask in plain language
Talk to it the way you'd talk to a coach. It pulls the right slice of your training to answer.
Review and act
Read the answer, and if it proposes plan changes, accept, edit, or decline each one. You stay in charge.
Straight answers.
Which AI assistants work?
Any assistant that supports MCP — the Model Context Protocol — including Claude and ChatGPT. You connect Eccentric once and ask away.
Does it change my training on its own?
No. Your assistant can read your data and draft proposed plan changes, but nothing is applied until you review it and accept. You're always the one who decides.
Is my data safe?
The connection is yours to grant and revoke, and your assistant only reaches your training when you allow it. Your data stays yours — and you can export all of it as a CSV whenever you like.
Do I need to be technical to set this up?
No. Connecting is a one-time step in your assistant, and from then on you just talk to it in plain language.