CoachLuca

This workspace is the profile-owned root for the Coach Luca multi-agent coaching system.

Purpose

Use this repository as the single source of truth for:

  • shared doctrine and operating rules
  • coordinator and specialist workspaces
  • athlete-scoped intake, planning, data, media, and delivery records
  • canonical athlete daily and weekly logs, plus any legacy export artifacts that still exist
  • session and long-term memory artifacts

Active Layout

  • athlete/: canonical athlete roots, per-athlete settings, delivery config, official notes, athlete data, media, and athlete-scoped specialist records
  • shared/: doctrine, ownership rules, templates, and automation notes
  • planning/: proposal, implementation plan, and review artifacts
  • intake/: shared intake workflow documentation only, not athlete state
  • team/coach-luca/: coordinator bootstrap files, synthesis workspace, and local logs
  • team/endurance/: endurance specialist bootstrap and working files
  • team/nutrition/: nutrition specialist bootstrap and nutrition working files
  • team/recovery/: recovery specialist bootstrap and working files
  • team/strength/: strength specialist bootstrap and working files
  • memory/: dated session notes and historical context
  • state/: runtime and system state artifacts

Operating Rules

  1. Keep every coaching and athlete artifact inside this workspace root.
  2. Treat shared/doctrine/ as the cross-agent policy layer.
  3. Treat athlete/ as the canonical root for athlete-specific planning, settings, delivery config, outputs, intake, data, and media.
  4. Treat athlete/logs/daily/ and athlete/logs/weekly/ as the official human-readable output locations.
  5. Treat athlete/athlete-notes/ as a legacy export or archive location when it still exists. Do not create or rely on athlete/team-reports/ for current operations.
  6. Keep specialist reasoning in specialist-owned workspaces and in the canonical athlete daily-log sections.

Status

Phase-one scaffolding is implemented. Manual dry-run and quality review are still required before any broader automation is considered complete.