complete_program
ChatGPTMark the user's active Aptiva workout program as completed once every day is finished. Fails and lists the unfinished days otherwise.
create_custom_workout
ChatGPTBuild a NEW custom workout for the authenticated user from ONLY real, category-scoped exercises, fitted to a target duration, and persist it via ms-user-workouts. Every generated workout ENDS with a short dedicated "Stretching" cool-down section (up to 3 stretches, max 2 minutes) whose time is INCLUDED in target_duration_minutes — do not ask for extra time for it, and do not tell the user to stretch separately; the result reports stretching_duration and stretching_count. If the scoped exercise pool is too small to fill the target, exercises are repeated as extra rounds (max 3 passes) and then per-exercise time is increased (up to 2x); the result reports target_met and duration_shortfall_seconds — relay the real duration to the user when the target is not met.
create_workout_program
ChatGPTCreate the authenticated user's PRIVATE workout program (user-scoped in ms-user-workouts, NOT the public catalogue). References existing catalogue workouts by their IDs.
generate_meal_plan
ChatGPTGenerate a new daily meal plan (calorie and macro targets plus Breakfast, Lunch, Snack and Dinner) for today, a past day or up to 7 days ahead. Use when the day has no plan or the user asks for a new one. Meals already eaten are kept. If a plan exists the tool asks for confirmation first; call again with confirm_replace=true only after the user agrees. Requires Premium. Takes about 10 seconds.
generate_workout_program
ChatGPTStart generating a new personalised Aptiva workout program for the user from their fitness profile (and their last completed program, if any). Generation runs in the background and takes about a minute; afterwards call get_program_generation to check on it and activate it. Requires Premium. Refuses when the user already has an active program. Optional feedback (how the last program went, what to change) is saved on the completed program and steers the new one. Calling it again while a generation is running returns the same job.
get_activity
ChatGPTSummarise the user's recent Aptiva activity: water intake, mood check-ins, weight log and completed program days. Choose period 'today', '7d' (default) or '30d', or pass from/to dates (YYYY-MM-DD) to override.
get_my_workout_program
ChatGPTReturn the authenticated user's PRIVATE active workout program and schedule (user-scoped in ms-user-workouts). This is NOT the public catalogue: use list_all_workouts to browse public workouts.
get_nutrition_today
ChatGPTRead the user's live nutrition for a day in Aptiva: calorie and macro targets, planned meals, what is logged and what is left. Use for "what did I eat", "what's left today" or before suggesting what to eat next. Works for today, a past day, or up to 7 days ahead. The reply always states the date.
get_profile
ChatGPTReturn the authenticated user's stable id, name (when known) and email
get_program_generation
ChatGPTCheck on a program started with generate_workout_program. While it is still generating returns status pending; once ready it activates the program as the user's current program and returns it (status complete). Returns failed if generation failed (start again), blocked if another program became active meanwhile, and none if there is nothing generating. job_id defaults to the most recent generation.
get_program_status
ChatGPTShow where the user is in their active Aptiva workout program: which Day N is next, which days are completed, in progress or skipped, with titles, durations and links to open each day in Aptiva. Use it to answer 'what is my workout today?' or 'how is my program going?'. Returns program: null when the user has no active program.
get_user_fitness_profile
ChatGPTRetrieve the authenticated user's fitness profile: goals, level, body metrics and this week's calories burned (kcal_week)
list_all_exercises
ChatGPTList all exercises from the database with complete details
list_all_workouts
ChatGPTList all workouts from the database with complete details
list_my_workouts
ChatGPTList the authenticated user's own saved custom workouts (user-scoped in ms-user-workouts). Not the public catalogue: use list_all_workouts for that.
log_food
ChatGPTLog a free-text food the user ate, for a meal slot, today or a past day only. The description MUST include quantities (e.g. "two eggs and a slice of toast"); if portions or the slot are unclear, ask first, never guess. Calories and macros are estimated. If it is the dish already planned for that slot, use log_planned_meal instead. A real slot replaces that slot's unlogged planned meal; use "Added" for an extra.
log_hydration
ChatGPTRecord water the user drank, in millilitres, in Aptiva. Optional 'at' is an ISO datetime, default now.
log_mood
ChatGPTRecord how the user feels in Aptiva's mood tracker, with an optional short note. Optional 'at' is an ISO datetime, default now.
log_planned_meal
ChatGPTMark a meal already on the user's plan as eaten, today or a past day only, when they ate the planned dish. Identify it by slot (Breakfast/Lunch/Snack/Dinner) and/or its name from get_nutrition_today.
log_weight
ChatGPTRecord the user's body weight in kilograms in Aptiva (also updates their current weight). Use when the user says they weighed in. Optional 'at' is an ISO datetime, default now.
modify_workout
ChatGPTModify an existing catalog or custom workout (shorten/extend/swap or replace a muscle group) using only category-scoped exercises. Catalog sources are copied-on-write; custom sources are patched in place. Optionally repoint a program day. If the workout already ends with a "Stretching" cool-down section it is preserved exactly as-is and target_duration_minutes still means the TOTAL time including it (this tool never adds a cool-down — only create_custom_workout does); stretching_duration and stretching_count report what was carried over.
swap_meal
ChatGPTSwap a planned, not yet eaten meal for a different macro-similar dish in the same slot, for today, a past day or up to 7 days ahead. Identify it by slot and/or name. Pass disliked ingredients as avoid (e.g. ["chicken"]) and the swap tries to avoid them (at most two swaps per call).
update_coach_preferences
ChatGPTSave preferences for the user's Aptiva AI coach (for example tone or reminders) as simple key/value pairs. New values are merged into the existing preferences. Returns the merged preferences.
update_program_day
ChatGPTReplace the workout on one day of the user's active Aptiva program with a workout from the catalogue (use list_all_workouts for ids). Days that were already started cannot be replaced.