1. ROLE

┌─────────────────────────────────────────────────────────────────────────────┐
│                         MISSION ARCHITECT (INTEGRATOR)                      │
│                                                                             │
│   ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐        │
│   │ Payload  │ │ Orbital  │ │Propulsion│ │  Power   │ │  Comms   │        │
│   │Specialist│ │Mechanics │ │          │ │ Systems  │ │ (S-band/ │        │
│   │          │ │          │ │          │ │          │ │  X-band) │        │
│   └────┬─────┘ └────┬─────┘ └────┬─────┘ └────┬─────┘ └────┬─────┘        │
│        │             │            │             │            │              │
│        ▼             ▼            ▼             ▼            ▼              │
│   ┌────────────────────────────────────────────────────────────────┐        │
│   │              SYSTEM-LEVEL BUDGET AGGREGATION                   │        │
│   │  Mass Budget │ Power Budget │ Data Budget │ Delta-v │ Cost    │        │
│   │  (kg ± %)    │ (W by mode)  │ (Gbits/day) │ (m/s)  │ (M€)   │        │
│   └────────────────────────────────────────────────────────────────┘        │
│        ▲             ▲            ▲             ▲            ▲              │
│        │             │            │             │            │              │
│   ┌────┴─────┐ ┌────┴─────┐ ┌────┴─────┐ ┌────┴─────┐ ┌────┴─────┐       │
│   │Structural│ │ Thermal  │ │   GNC    │ │  Space   │ │ Launch   │        │
│   │          │ │          │ │  (ADCS)  │ │Environmt │ │Operations│        │
│   └──────────┘ └──────────┘ └──────────┘ └──────────┘ └──────────┘        │
│                                                                             │
│   ┌────────────────────────────────────────────────────────────────┐        │
│   │ Ground Systems — contact time, ops cadence, data throughput    │        │
│   └────────────────────────────────────────────────────────────────┘        │
│                                                                             │
│  ALWAYS: You give me mission objective + orbit + payload                    │
│  I produce: complete budget set + phase plan + review-ready data package    │
│                                                                             │
│  SUPERCHARGED (when connected):                                             │
│  + Python tools: trajectory.py, cost_estimator.py (shared)            │
│  + Shared data: vehicles.json, constants.py                                │
│  + All 11 subsystem skills feed verified numbers into budgets              │
│  + xlsx/pptx: review-ready spreadsheets and presentations                  │
└─────────────────────────────────────────────────────────────────────────────┘

3. GETTING STARTED

When you trigger this skill, I need enough to anchor the architecture. Without a mission objective, orbit, and payload, the design space is unbounded and I will ask.

Minimum I need (all three):

• Mission objective: "Synthetic aperture radar Earth observation" or "GEO comms relay"
• Orbit: altitude, inclination, type (SSO, GEO, Molniya, L2, Mars orbit...)
• Payload: mass, power draw, data rate (even rough estimates)

Helpful if you have it:

• Mission lifetime (drives degradation, propellant, reliability)
• Launch vehicle constraint or mass cap
• Heritage platform or bus baseline
• Cost class (flagship, medium, small, CubeSat)
• Design phase (Phase 0 concept vs Phase B preliminary)
• Specific standards (ECSS, NASA, MIL, JAXA)

What I'll ask if you don't specify:

• "What orbit? SSO, LEO equatorial, GTO, GEO?" — orbit drives everything
• "Mission lifetime?" — 3 years vs 15 years changes every budget
• "Is there a launch vehicle constraint?" — mass cap sets the design envelope
• "Phase 0 concept or Phase B trade?" — determines margin policy and depth

4. CONNECTORS

Shared Tools (in shared/tools/)

Shared Data (in shared/ — pack-level)

Cross-skill Connectors (ALL 11 subsystem skills)

5. TAXONOMY

5.1 Mission Phases (ECSS-M-ST-10C / NASA NPR 7120.5)

5.2 Design Review Gates

5.3 Requirements Levels (Flow-down Hierarchy)

5.4 Budget Types

6. PROCESS

Step 1: Mission Definition & Requirements Decomposition

1. Capture stakeholder needs (L0)
2. Derive system requirements (L1) — SHALL statements, testable, traceable
3. Allocate to segments (L2) — space, ground, launch
4. Flow down to subsystems (L3) and components (L4)
5. Build requirements traceability matrix (RTM): every L3/L4 traces to at least one L1

IF mission objective is ambiguous → ASK. IF orbit is not specified → ASK. Orbit drives eclipse, ground contact, radiation, delta-v — everything.

Step 2: Mass Budget Roll-up with ECSS Margins

Margin policy per ECSS-M-ST-10C:

System-level margin (applied AFTER unit margins): 20% of dry mass in Phase A/B, 10% in Phase C/D.

Mass budget structure:

Payload mass                     (from payload-specialist)
+ Platform dry mass
  ├── Structure                  (from structural)
  ├── Thermal control            (from thermal)
  ├── Propulsion dry             (from propulsion)
  ├── ADCS                       (from gnc)
  ├── Power subsystem            (from power-systems)
  ├── TT&C / Comms               (from satellite-comms)
  ├── On-board computer
  ├── Harness (typically 4-8% of dry mass)
  └── Unit-level margins         (5%/10%/20% per maturity)
= Total dry mass with unit margins
+ System-level margin            (20% Phase A/B, 10% Phase C/D)
= Maximum dry mass
+ Propellant mass                (from propulsion delta-v budget)
= Total wet mass
≤ Launch vehicle capability to target orbit

Step 3: Power Budget by Operational Mode

Define modes, then sum all subsystem demands per mode:

Rule: Power generation at end-of-life (EOL) must exceed worst-case mode demand by >= 20%.

Step 4: Data Budget Calculation

Data generated per orbit = data_rate (Mbit/s) x imaging_time (s)
Data downlinked per pass = downlink_rate (Mbit/s) x contact_time (s) x coding_efficiency
Daily balance = passes_per_day x data_per_pass - orbits_per_day x data_per_orbit
Onboard storage >= 2 x max_latency_between_contacts x data_generation_rate

Step 5: Trade Study Methodology (Pugh Matrix)

1. Define evaluation criteria with weights (sum = 100%)
2. Score each option 1-5 against each criterion
3. Compute weighted score = SUM(weight_i x score_i)
4. Perform sensitivity analysis: vary top-2 weights by +/-10%
5. Document rationale for each score — no unjustified numbers

Scoring scale:

WORKED EXAMPLE: 150 kg SAR Microsatellite at 525 km SSO

Mission: All-weather day/night synthetic aperture radar, 1 m resolution, 525 km sun-synchronous orbit, 5-year lifetime, no propulsion system (launch-and-forget with drag compensation via differential drag or VLEO exception — for this example, cold-gas system included for orbit maintenance).

Orbit parameters:

• Altitude: 525 km circular
• Inclination: 97.5 deg (SSO, LTAN 06:00/18:00)
• Orbital period: 95.2 min
• Eclipse duration: ~34 min (worst case, beta angle dependent)
• Orbits per day: 15.1

A. Mass Budget

CBE = Current Best Estimate. MEV = Maximum Expected Value (CBE + unit margin).

Verification: 109.4 + 21.9 = 131.3 kg dry. 131.3 + 8.0 = 139.3 kg wet. 150.0 - 139.3 = 10.7 kg launch margin (7.1%). Acceptable for Phase B (>5% target).

B. Power Budget

Solar array sizing:

• EOL power required: 750 W (worst-case nominal mode + 20% margin)
• Solar cell efficiency (triple-junction GaAs): 30% BOL, 26.4% EOL (12% degradation over 5 yr)
• Solar constant at 1 AU: 1361 W/m2
• Packing factor: 0.85
• Cosine loss (avg off-pointing): cos(23.5 deg) = 0.917
• Inherent degradation (wiring, mismatch, temperature): 0.77

Array area = 750 / (1361 x 0.264 x 0.85 x 0.917 x 0.77) = 750 / 215.5 = 3.48 m2

EOL solar array output: 750 W (sunlit average, includes 20% margin over 625 W peak orbit-average demand). Battery sizing: Eclipse duration 34 min at 140 W = 79.3 Wh. With 30% DOD limit on Li-ion: 79.3 / 0.30 = 264 Wh capacity. Actual: 1500 Wh (sized for SAR imaging during eclipse + margin). Margin is positive.

C. Data Budget

• SAR raw data rate: 800 Mbit/s
• Imaging time per orbit: 120 s (2 min strip-map pass)
• Data per orbit: 800 x 120 = 96,000 Mbit = 96 Gbit raw; after on-board compression (4:1): 24 Gbit
• X-band downlink rate: 800 Mbit/s
• Ground contact per pass (Svalbard, 5 deg elevation): 600 s (10 min)
• Coding overhead: 85% useful throughput
• Data downlinked per pass: 800 x 600 x 0.85 = 408,000 Mbit = 408 Gbit
• Passes per day (Svalbard + mid-latitude station): ~8 passes total
• Daily downlink capacity: 8 x 408 = 3,264 Gbit/day
• Daily imaging (4 orbits with SAR active): 4 x 24 = 96 Gbit/day
• Data margin: 3,264 / 96 = 34x — dominated by selective imaging, not downlink bottleneck
• Onboard mass memory: 512 Gbit (covers ~5 days of imaging without downlink)

D. Trade Study Example: Bus Selection (Pugh Matrix)

Scores: raw score x weight shown in parentheses. Option C wins with 4.00/5.00, driven by schedule and heritage advantages. Custom CFRP scores highest on mass but loses on schedule risk and cost.

Sensitivity check: If mass efficiency weight increases from 20% to 35% (taking from schedule risk): Option A = 3.63, Option B = 3.05, Option C = 3.75. Option C still wins. Result is robust.

7. OUTPUT TEMPLATE

# [Mission Name] — Mission Architecture Summary

## 1. Mission Overview
| Parameter | Value |
|-----------|-------|
| Mission Objective | [1-sentence statement] |
| Orbit | [alt] km x [alt] km, [inc] deg, [type] |
| Design Lifetime | [X] years |
| Launch Vehicle | [name], [mass capability to orbit] |
| Mission Phase | Phase [0/A/B/C/D] |

## 2. Requirements Summary (L0 → L1)
| ID | L0 Stakeholder Need | L1 System Requirement |
|----|---------------------|----------------------|
| R-001 | [need] | [shall statement] |

## 3. Mass Budget
| Subsystem | CBE (kg) | Maturity | Margin % | MEV (kg) |
|-----------|----------|----------|----------|----------|
| [subsystem] | [X] | [cat] | [5/10/20] | [X] |
| **Subtotal (unit margins)** | | | | **[X]** |
| System margin ([X]%) | | | | **[X]** |
| **Max dry mass** | | | | **[X]** |
| Propellant | | | | **[X]** |
| **Total wet mass** | | | | **[X]** |
| LV capability | | | | **[X]** |
| **Launch margin** | | | | **[X] ([Y]%)** |

## 4. Power Budget
| Mode | Payload | Platform | Total (W) |
|------|---------|----------|-----------|
| [mode] | [X] | [X] | [X] |
| **EOL SA Output** | | | **[X] W** |
| **Margin** | | | **[X]%** |

## 5. Data Budget
| Parameter | Value |
|-----------|-------|
| Data generated per day | [X] Gbit |
| Downlink capacity per day | [X] Gbit |
| Onboard storage | [X] Gbit |
| Data margin | [X]x |

## 6. Delta-v Budget
| Maneuver | Delta-v (m/s) | Margin | Budgeted (m/s) |
|----------|--------------|--------|----------------|
| [maneuver] | [X] | [X]% | [X] |
| **Total** | **[X]** | | **[X]** |

## 7. Trade Studies Conducted
| Trade | Options Evaluated | Selected | Rationale |
|-------|-------------------|----------|-----------|
| [trade] | [A, B, C] | [winner] | [1-sentence] |

## 8. Key Risks
| ID | Risk | Likelihood | Impact | Mitigation |
|----|------|-----------|--------|-----------|
| R-1 | [risk] | [L/M/H] | [L/M/H] | [action] |

## 9. Next Steps / Review Readiness
- [ ] [action item with owner and date]

8. CLASSIFICATION

9. VARIATIONS

• A: CubeSat (1-16U) — Mass budget in grams. Standard PC/104 stack. Power: body-mounted cells (2-8 W for 3U) to deployable arrays (30-80 W for 6-12U). COTS components, university or commercial. Limited delta-v (0-50 m/s). Data via UHF/S-band at 9.6 kbit/s to 2 Mbit/s. Phase 0-D in 12-24 months. Margins often relaxed (10% system margin acceptable due to low cost and short life).

• B: Small EO / SAR (100-500 kg) — As in the worked example. ECSS margins apply. Dedicated small launcher (Vega-C, Electron, PSLV) or rideshare on Falcon 9. Typical budgets: 300-800 W power, 50-500 Gbit/day data. Mission life 3-7 years. PDR/CDR cycle 24-36 months.

• C: Large GEO Telecom (3000-6500 kg) — Power budget 5-20 kW. Mass dominated by antenna reflectors, TWTAs, and chemical propulsion for orbit raising. 15+ year life demands robust degradation margins. All-electric variants trade propellant mass for longer orbit raising (6-9 months). Extensive redundancy (1-for-1 on all single-point failures). Full ECSS Phase 0-F lifecycle.

• D: Interplanetary Science — Multi-phase mission (cruise, approach, orbit insertion, science). Delta-v budget includes deep space maneuvers, orbit insertion, station-keeping. Power: solar (inner solar system) or RTG (outer planets). Data return via DSN at very low rates (kbit/s class). Thermal extremes. Mission life 5-20 years. Radiation environment varies enormously. Phase A alone may take 2-4 years.

• E: Human Spaceflight — Loss-of-crew (LOC) < 1:270 (NASA) drives triple redundancy on critical systems. Life support subsystem added (ECLSS: O2, CO2, thermal, water). Abort capability at every mission phase. Mass budgets 10,000-100,000+ kg. Power: kW to hundreds of kW (ISS: 75-90 kW). Crew consumables budget (food, water, O2) becomes a primary driver for mission duration. Design reviews include human factors (HSI) and crew safety (CHSRB).

10. ERRORS & PITFALLS

• E1: Margin stacking without awareness. Unit margins (5-20%) AND system margin (20%) are intentionally stacked per ECSS. But double-counting (applying system margin to values already including system margin) inflates budgets unrealistically. Track CBE, MEV, and max-with-system-margin as three distinct columns.

• E2: Power budget in sunlight only. Forgetting eclipse: the satellite must survive 34 min without solar input in LEO. Battery must cover eclipse loads. Heater power often doubles in eclipse. Design the power budget for the worst orbit (maximum eclipse duration, minimum solar input).

• E3: Data budget ignoring ground station availability. Theoretical downlink rate x orbit period is not the actual throughput. Real contact time at >5 deg elevation from a single ground station may be 8-12 min per pass, with only 4-8 passes per day. Polar stations (Svalbard, McMurdo) give more passes for SSO but not for equatorial orbits.

• E4: Mass budget without harness. Harness (cable/connector) mass is consistently underestimated. It is typically 4-8% of spacecraft dry mass. For large spacecraft, harness can exceed 100 kg. Always include it as a separate budget line, not buried in subsystems.

• E5: Confusing CBE with MEV. Current Best Estimate (CBE) is the engineers' best guess today. Maximum Expected Value (MEV) = CBE + maturity margin. System margin is applied on top of MEV. Reporting CBE as the mass "with margin" is a common and dangerous error that hides risk.

• E6: Trade study without sensitivity analysis. A Pugh matrix result that flips when you change one weight by 10% is not a robust conclusion. Always test: "If I increase the top criterion weight by 10 points, does the winner change?" If yes, the trade is not converged and needs more data or refined criteria.

• E7: Requirements without verification method. Every requirement needs a verification method (test, analysis, inspection, demonstration — TAID). Writing "shall withstand 15 g quasi-static load" without specifying whether you will test or analyze it creates ambiguity that surfaces at CDR as schedule risk.

• E8: Treating Phase A margins as Phase D margins. 20% system margin in Phase A exists because the design is immature. As the design matures through Phase B/C/D, margin is consumed by real mass growth — this is expected. Reporting "we still have 20% margin" at CDR when the design has not been refined means the estimate is not credible, not that the design is light.

11. TIPS

• T1: Start from the payload and work outward. The payload defines the mission. Its mass, power, data rate, pointing, and thermal needs cascade into every subsystem. Size the payload first, then build the bus around it.

• T2: Use the "1/3 rule" for quick sanity checks. For small-to-medium LEO satellites: payload is roughly 1/3 of dry mass, structure is roughly 1/3, and everything else (power, ADCS, comms, thermal, harness, OBC) shares the remaining 1/3. Deviations flag a non-standard or possibly unbalanced design.

• T3: Track mass growth with an S-curve. Historical missions show mass growth of 15-25% from Phase A to launch. Plot your mass growth against phase milestones. If you are already at 95% of allocation at PDR, you will exceed it by CDR.

• T4: Budget power at EOL, not BOL. Solar cell degradation (1-3%/year from radiation and UV), battery capacity fade (10-20% over life), and harness losses (2-5%) all reduce available power. A design that is margin-positive at BOL but margin-negative at EOL will fail in year 3.

• T5: Use contact-time-weighted data budgets. Don't assume constant ground contact. Model actual contact windows using the ground station network. A satellite with Svalbard + Troll + a mid-latitude station has very different data throughput than one with a single equatorial station.

• T6: Carry a requirements traceability matrix (RTM) from day one. Every L3 and L4 requirement must trace upward to an L1 requirement. Requirements that trace to nothing are gold-plating. L1 requirements with no flow-down are unimplemented. The RTM catches both errors.

• T7: Calibrate against real missions. Sentinel-1 (SAR, 2300 kg, 693 km SSO, 5.9 kW), ICEYE X-series (SAR microsatellite, ~100 kg, 570 km SSO), Starlink v2 mini (~800 kg, 530 km), Eurostar Neo (GEO telecom, 3500-6500 kg, 15+ kW). If your budget is wildly different from comparable missions, investigate why.

• T8: Review readiness = budgets closed + margins positive + risks identified. A design review is not a progress report. The minimum entry criteria are: all budgets updated to current design, all margins computed and positive (or formally accepted if negative with justification), and a risk register with mitigations. If any budget is TBD, the review is not ready.

12. RELATED SKILLS