import { ActivitiesSummary, ActivitiesSummaryOptions, ActivityDetailSummary, ActivityDetailSummaryOptions, TrainingOverview, TrainingOverviewOptions, WellnessOverview, WellnessOverviewOptions } from '../types';
import { ModuleConstructor } from './types';
export declare function applyCoachModule(Base: ModuleConstructor): {
    new (...args: any[]): {
        /**
         * Build the long-term training background intended for AI coach prompts
         * and future MCP responses. This method always queries running and
         * cycling sources together, validates the token before fan-out, and
         * returns a stable `training_overview_v1` shape instead of Garmin's
         * large raw payloads.
         *
         * Data included:
         * - athlete profile: age, gender, height, weight, observed sports
         * - current training state: status, load trend, acute/chronic load, ACWR
         * - load balance: low aerobic, high aerobic, anaerobic values vs targets
         * - sport summaries: run/bike volume, key form metrics, VO2, threshold,
         *   race prediction, FTP/power-to-weight, cycling ability
         * - multi-sport interpretation, training status trend, data completeness
         *   flags, and AI hint tags
         */
        getTrainingOverview(options?: TrainingOverviewOptions): Promise<TrainingOverview>;
        /**
         * Build the recovery-state overview intended for daily AI coach
         * readiness prompts. It normalizes Garmin HRV, sleep, body battery,
         * resting heart rate, respiration, SpO2, and skin temperature signals
         * into a stable `wellness_overview_v1` shape.
         *
         * Users often have incomplete wellness data; unavailable metrics are
         * marked as `no_data` or `partial` rather than omitted.
         */
        getWellnessOverview(options?: WellnessOverviewOptions): Promise<WellnessOverview>;
        /**
         * Build a low-token, high-coverage summary of what recent training
         * happened. Use this by default when an AI coach needs activity context
         * without loading single-activity details.
         *
         * The response uses `activities_summary_v1` and removes Garmin's large
         * raw list payload fields such as privacy, owner, map, media, and
         * verbose split metadata. It keeps the data useful for training advice:
         * sport counts, duration, distance, elevation, load, heart-rate zones,
         * training effect, running form, body-battery delta, and computed flags.
         */
        getActivitiesSummary(options?: ActivitiesSummaryOptions): Promise<ActivitiesSummary>;
        /**
         * Build a compact single-activity detail for AI coach analysis. Load
         * this only when the AI coach needs to analyze one specific workout
         * with finer metrics.
         *
         * The response uses `activity_detail_v1` and keeps the key details for
         * deeper post-workout reasoning: summary metrics, training impact,
         * intensity, run form, stamina, subjective feedback, sensor presence,
         * split availability, and AI hint tags.
         */
        getActivityDetailSummary(options: ActivityDetailSummaryOptions | number): Promise<ActivityDetailSummary>;
        client: import("../../common/HttpClient").HttpClient;
        url: import("../UrlClass").UrlClass;
        domain: import("../types").GarminDomain;
        getUserProfile(): Promise<import("../types").ISocialProfile>;
    };
};