Accountability

How the escrow works and what happens if things go wrong

Q: What specifically are the ~100 on-chain events per hectare per year?

The 100 events/ha/year is a planning estimate reflecting expected average farming activity:

• ~50 VeData batch anchors — cultivation logs, Bio-ID updates, harvest events (~weekly during growing season)
• ~25 labor contract engagements — 2 permanent workers × 12 monthly payments + 1–2 harvest khoán contracts per year (Source: CropTex; QĐ 3763/QĐ-UBND Đắk Lắk 19/12/2019)
• ~5 commercial escrow transactions — advance deposit + final settlement per harvest batch
• ~8 one-time setup events — DID registration, Farm NFT minting, harvest module activation

Path B audit decommits (~2,190/year) are a fixed system cost that does not scale per hectare and is excluded from this formula. Active farms during peak harvest months will generate more; smaller or less active farms will generate fewer. The ~1,200,000 total is the upper bound planning estimate.

---

Q: Is "25 labor contracts/ha/year" based on actual Đắk Lắk employment data?

Yes, derived from official field data. From QĐ 3763/QĐ-UBND Đắk Lắk (19/12/2019) and CropTex field research: 1 ha durian requires 2 permanent workers year-round at 5–7M VND/person/month. 2 workers × 12 monthly payments = 24 labor payment engagements/year, plus 1–2 harvest khoán contracts per harvest season. Total: ~25–26 individual labor engagements/year. The GA uses 25 as a conservative lower bound.

---

Q: Is "5 commercial escrow tx/ha/year" consistent with actual durian trade structure?

Yes. Đắk Lắk durian trade typically involves: 1–2 advance deposits from traders before harvest + 1–2 final settlements after delivery + occasional partial payments in multi-batch sales. ~3–5 commercial transactions/ha/year is the realistic estimate. The GA uses ~5 as the planning upper bound. (Source: DDC Holdings field experience across 46 communes; confirmed by trade structure analysis from the 2025 scandal investigation)

---

Q: Under a realistic downward scenario (~700,000 events/year), are milestones still achievable?

Yes. M7 requires ≥100,000 on-chain transactions during the M7 period (1–2 months). At 700,000 events/year, 2 months = ~117,000 transactions — M7 passes comfortably. The 700,000 scenario is likely the Year 1 reality as the ecosystem builds up, not a failure scenario. It becomes the baseline from which organic growth occurs as more enterprises integrate the SDK.

---

Q: How does the design prevent gaming M7's "≥100,000 on-chain tx" acceptance criterion?

Four independent layers: (1) Each transaction costs real MAGIC credits paid by users — generating empty transactions is economically self-defeating. (2) 5-layer VeData validation rejects invalid data before it reaches Cardano; fake records cannot pass biological plausibility checks. (3) IOB Member 3 independently verifies field evidence — GPS photos, district reports, and dashboard data must corroborate on-chain activity. (4) Random 5% spot checks require field inspectors to confirm that on-chain records correspond to physical farms and trees.