Morpheus App (Joel Jamieson) — User Guide PDF

The Morpheus app is the primary UI reference for the user — they want Parasym to "replicate a little bit" of Morpheus but with HealthKit openness.

Key Morpheus design patterns (verified from the PDF + support articles):

1. Recovery Score, 0-100% with 3-color coding:

• 🔴 Red: 0-40% (Strained / Run down)
• 🟡 Amber: 41-80% (Balanced)
• 🟢 Green: 81-100% (Primed)
• Color zones are dynamic — change daily based on HRV vs personal 10-day rolling baseline

1. Recovery Score factors (ranked by influence, per the Morpheus user guide):

• HRV (biggest factor) — rolling 10-day average, 1σ = range of ~65-80%
• Sleep amount + quality
• Previous day's training (post-workout recovery adjustment)
• Steps (smaller role)
• Subjective fatigue markers
• With only 3 HRV measurements in 10 days → bigger range, more weight on other data
• Other variables can swing score 10-20%

1. Daily Recovery Test: ~2.5 min, 30+ R-R intervals sampled, sets the day's recovery score + training zones

1. Dynamic Heart Rate Zones — 3 zones, recomputed daily from the recovery score:

• Blue (Low Intensity / Recovery) — recovery-promoting
• Green (Moderate / Fitness) — improves conditioning
• Red (High Intensity) — upper limit, too much = recovery debt

1. Zone-Based Interval Training (ZBIT) — 12 interval templates, all driven by the dynamic zones

1. Weekly Targets — CardioSmart, recommends duration in each zone per week

1. Initial state: "100%" for the first 5-7 days until baseline establishes. Discontinuing for 7-10 days requires re-establishing baseline. Best tested in morning before caffeine. Same position each time (lying down or seated).

Pattern synthesis (per the user's goal)

The user said: "my goal is to build something like [NOOP] and [goose] but obviously built for people with HRV chest straps (morpheus mainly, but should also be capiable to be used with the polar H10 and other chest straps). collating the data from both apple health."

Reading the three projects together:

| Aspect | NOOP | Goose | Morpheus | Parasym (target) | |---|---|---|---|---| | Strap target | WHOOP 4.0/5.0 (proprietary protocol) | WHOOP 5.0 (proprietary) | Morpheus M7 (BLE standard + custom) | M7 + Polar H10 + any BLE HRM (standard GATT) | | HealthKit | Imports WHOOP + Apple Health history | Not in v0.1 | Closed system, doesn't sync to Apple Health | Read AND write (the differentiator) | | Recovery score | Yes (Plews/Buchheit + Lamberts + Gabbett + Foster) | Daily recovery view | Yes (HRV-set, rolling 10-day, dynamic zones) | Yes (Galpin σ-deviation rule per brain chunks 6472/6475) | | Dynamic training zones | Implied | Implied | Yes (3 zones, daily) | Yes (3 zones, daily) | | Data storage | Local SQLite | Local + Rust core | Cloud (closed) | Local SwiftData + optional iCloud (v0.3+) | | Agent / AI layer | None | None | None | Yes — TARS integration (v0.3+) | | License | PolyForm Noncommercial 1.0.0 | Alpha proof of concept | Closed / paid | MIT | | Local-first | ✅ | ✅ | ❌ | ✅ |

The Parasym niche:

• Like NOOP/Goose: local-first, no cloud, no account, no subscription, open source
• Like Morpheus: recovery score + dynamic training zones + chest strap primary
• UNLIKE all three: HealthKit read AND write (Morpheus is the only closed one; NOOP imports but doesn't write to Apple Health)
• UNLIKE all three: agent-aware (the v0.3 TARS integration is the long-term play)

Design questions for the user (direction sign-off needed)

The user said "lets agree on this direction first." Surfacing the questions that the research surfaced, in priority order:

A. Morpheus zone UI — replicate or invent?

Morpheus uses 3 zones (Blue/Green/Red) with daily recomputation. The user explicitly said "this could also be interesting to try and steal and replicate inside parasym."

Recommendation: replicate the 3-zone model. It's tested, it's intuitive, it's a UI pattern users already understand. But make the colors + thresholds configurable (the "what's Red for me" slider).

Alternative: 5-zone model (Recovery / Z1 / Z2 / Z3-4 / Z5-VO2 max) per the standard TrainingPeaks / Galpin / Peter Attia framework. More precision, less accessible. Adds complexity.

B. Recovery score formula — Morpheus vs Galpin σ-deviation

• Morpheus: HRV sets the RANGE (1σ = 65-80%), then other factors slide within that range. Multi-factor.
• Galpin σ-deviation rule (already in gbrain, verified from FoundMyFitness): 1σ = yellow flag, 2σ = "something's happening." Single-factor (just HRV). Simpler, more transparent, more defensible.

Recommendation: hybrid. Use Galpin's σ-deviation for the PRIMARY signal (it's the public-domain, single-factor, transparent), but surface the secondary signals (sleep, training load, RHR) as modifier insights. The daily score is RMSSD-driven; the cross-source correlation insights are the "what changed" story.

C. NOOP-style "Readiness" composite vs Morpheus-style recovery-only

NOOP's Readiness synthesizes 4-5 published frameworks (Plews/Buchheit, Lamberts, Gabbett, Foster). More comprehensive, more defensible. Higher complexity.

Recommendation: ship Morpheus-style recovery + Galpin σ-deviation in v0.1. Add the NOOP-style multi-factor synthesis in v0.3+ if the user wants the depth. v0.1 ships working with one formula; v0.3 enriches with multi-factor.

D. Strap-agnostic — how far to go?

The user said "morpheus mainly, but should also be capiable to be used with the polar H10 and other chest straps." This is the same scope as the v0.1 plan (BLE HR Service, any HRM exposing R-R intervals).

Recommendation: keep the v0.1 scope. BLE standard is the abstraction. M7 + Polar H10 are the two test devices. Future straps drop in if they follow the spec.

E. Agent / AI layer — now or later?

The user wants something "like" NOOP/Goose. The GOOSE part (Block/goose, the agent framework) is different from b-nnett/goose (the iOS app). Per the conversation context, the user means the agentic / AI-aware pattern: Parasym's data flows to TARS (or another agent) for context-aware conversation.

Recommendation: defer to v0.3+ per the existing concepts/parasym-tars-integration design. v0.1 ships the data foundation + insight engine. v0.3 ships the TARS integration (Option 1: file-based daily summary → gbrain).

F. UI feedback the user gave (chair padding, top padding)

The screenshot shows the DashboardView too close to the navigation bar AND the SettingsSection padding is too tight. This is a small SwiftUI fix (the ScrollView/VStack padding needs more breathing room). Not a design question; just a v0.1.1 polish.

Linked concepts

• projects/parasym — parent project hub (rev 3 spec)
• parasym tars integration — v0.3+ integration design
• hrv sdnn vs rmssd — the metric mismatch Parasym resolves
• people/joel jamieson — Morpheus designer
• morpheus m7 — primary data source

Source

• NoopApp/noop GitHub README (researched 2026-06-22)
• b-nnett/goose GitHub README (researched 2026-06-22)
• Morpheus App User Guide PDF (researched 2026-06-22)
• Joel Jamieson, Peter Attia Drive #305 (Jun 2024) — RMSSD methodology
• Olav Aleksander Bu, Peter Attia Drive #294/331 — readiness philosophy
• Andy Galpin, FoundMyFitness — σ-deviation rule (gbrain chunks 6472/6475)