Generate with AI

You don't have to learn ReptoLang to use it. ReptoLang is plain text, which means an LLM — ChatGPT, Claude, Gemini, or whatever you already use — can write a whole program for you. You describe what you want, the model writes the .repto, and you compile it in the Playground.

How it works

Build your prompt. Fill in the form below — goal, days per week, experience, equipment — and we'll assemble a ready-to-paste prompt. Hit Copy.
Paste it into your LLM. Drop it into ChatGPT, Claude, or any assistant. The prompt links to these docs so the model can learn the syntax and the exercise names.
Run the result. Paste the program the model gives you into the Playground to compile it instantly and catch any issues.

Build your prompt

Fill in the form and copy the prompt it generates. You don't need to read or understand that prompt — it's written for the AI, not for you. It includes the full ReptoLang syntax inline, so it works even with assistants that can't open links (like the free tier of most chatbots).

Goal

Days / week

Experience

Length (weeks)

Unit

Equipment

Focus lifts (optional)

Notes (optional)

Your prompt

You are an expert strength coach who writes programs in ReptoLang, a plain-text language for strength programs.

ReptoLang quick reference — everything you need is here, no need to open any links:

Program structure
- Begin every program with: @version 1
- Optional metadata: @name "...", @author "...", @description "...",
  @difficulty beginner|intermediate|advanced|elite,
  @category powerlifting|bodybuilding|strength|hypertrophy|conditioning|general,
  @unit kg|lb|lbs
- Optional tm { } block defines training maxes that percentage sets reference:
    tm {
      Squat (Back Squat) = 140kg
      Bench Press = 100kg
    }
- The body is one or more weeks. Each week has days; each day has exercise blocks; each line inside
  an exercise block is one set:
    week "Week 1" {
      day "Day A" {
        Squat (Back Squat) {
          5 @%75
          5 @%85
        }
      }
    }
- Scaled-down week:  deload week "Deload" { scale: 0.9  use: week 1 }
- Superset (shared rest timer):
    superset {
      rest: 60s
      Barbell Curl { 3x10 @40kg }
      Tricep Pushdown { 3x12 @25kg }
    }

Set syntax —  [warmup] <reps> @<weight> [drop|failure]
- Reps:   5 (fixed) · 8-12 (rep range) · 3x8 (3 sets of 8) · amrap or 5+ (as many as possible) · 30s or 1m30s (time-based)
- Weight: @%85 (85% of this lift's TM) · @%85 of Squat (% of another lift's TM) · @135lbs or @60kg (fixed load) · @BW (bodyweight) · @BW+20kg (bodyweight plus load)
- Flags:  warmup (before the reps) · drop / failure (after the weight)
- Optional options before a block's sets:  rest: 3m · intensity: rpe|rir · notes: "..."
- // starts a line comment

Worked example
@version 1
@name "Upper / Lower Strength"
@difficulty intermediate
@category strength

tm {
  Bench Press = 100kg
  Squat (Back Squat) = 140kg
}

week "Week 1" {
  day "Upper" {
    Bench Press {
      warmup 5 @%50
      5 @%80
      amrap @%85
    }
    Pull-Ups {
      3x8 @BW
    }
  }
  day "Lower" {
    Squat (Back Squat) {
      5 @%75
      5 @%85
    }
  }
}

Now write a complete, valid ReptoLang program for me:
- Goal: General strength
- Training days per week: 4
- Experience level: Intermediate
- Program length: 8 weeks
- Equipment available: Barbell, Dumbbells
- Weight unit: lb

Rules:
1. Start with `@version 1`, then add @name / @difficulty / @category metadata.
2. Include a tm { } block for any percentage-based (@%) sets.
3. Refer to each exercise by a common, standard name (e.g. "Bench Press", "Squat (Back Squat)", "Deadlift (Conventional)", "Pull-Ups"). I'll verify the names when I compile it.
4. Use only the set syntax shown in the reference above.
5. Output ONLY the ReptoLang program in a single code block — no commentary.

Full documentation, if you're able to open links:
- Getting started: /docs/reptolang/getting-started
- Sets & syntax: /docs/reptolang/syntax
- Progressions & hooks: /docs/reptolang/progressions
- Language reference: /docs/reptolang/reference
- Exercise library: /docs/exercises

When you're done I'll paste the program into the ReptoLang Playground (/playground) to compile and check it.

You don't need to read this — it's written for the AI. Paste it into ChatGPT or Claude, then bring the program back to the Playground → to compile it.

Tips for good results

Be specific. The more detail you give in Focus lifts and Notes (session length, injuries, equipment quirks), the better the program fits you.
Use real exercise names. Ask the model to reference exercises by their exact names from the exercise library — that's what makes them resolve cleanly.
Always compile it. Paste the result into the Playground and check for warnings or unresolved exercises. If something looks off, paste the warning back to the LLM and ask it to fix it.
Iterate. Treat it like a conversation — "make week 4 a deload", "swap the machine work for free weights", "add an AMRAP set on the last bench day".

Once you're comfortable, the Getting started guide shows you how to read and tweak the program by hand.