Community research · Posted article

Project Research Dossier — Minara AI | Pre-TGE Snapshot — Sep 30, 2026

By sulloctea

Open original Discord thread

Bottom line: Minara is a shipped, audited AI trading product with a credible founder track record, but it has no published tokenomics. Its points ("Sparks") are sold alongside subscriptions, so the biggest long-game risk is how Sparks convert at TGE. Full PDF attached.

1. What it builds

  • AI "virtual CFO" for crypto: chat research, on-chain analytics and trade execution in one app. Autopilot runs strategies on USDC-margined perps via Hyperliquid or Lighter.
  • Open-source minara-cli (MIT): spot/perp trading, swaps, multi-wallet from the terminal.
  • Beta Aug 14, 2025; paid tiers live in 2026.

2. Constraint check

  • Token: none announced ✅
  • Public raise / ICO: none. Only private backing from Circle Ventures (amount undisclosed) ✅
  • Points: "Sparks" live, future conversion unconfirmed ⚠️

3. Tokenomics: known vs unknown FDV, supply, allocations, vesting, cliffs, emissions: all unpublished. Only hard data = Sparks bundled with annual plans:

  • Lite $16/mo → 1,920 Sparks
  • Starter $40/mo → 4,800 Sparks
  • Pro $165/mo → 19,800 Sparks = a flat ~10 Sparks per $1 paid on every tier. If Sparks convert pro-rata, this becomes a de-facto pre-sale without published terms. Value per Spark = (FDV × Sparks allocation) / total Sparks. The last two inputs are undisclosed, so dilution can't be sized yet.

4. Unlock & dilution (illustrative, NOT project data) Assume 10% Sparks allocation, 50% unlocked at TGE → 5% of supply held by users with a known ~$0.10/Spark cost basis. If implied value > $0.10, holders are in profit on day 1 → expect sell pressure. A single team/investor cliff at 6-12 months would be a second shock. TGE checklist: total supply & FDV · circulating % at TGE · allocation split (team/investors/treasury/Sparks) · cliff vs linear per bucket · total Sparks issued (bought vs earned) + snapshot date · emissions & any revenue buyback.

5. Team verification ✅ Verifiable: CEO Lowes Yang (co-founded NFTGo; Circle backed NFTGo in 2022) · public GitHub (MIT, 68 commits) · SlowMist audit Jul-Aug 2025: 1 critical, 2 high, 8 low, 12 suggestions; critical/high fixed. ⚠️ Marketing: "Trusted by" logos (Circle, Ripple, Hyperliquid, Lighter…) are partners, not confirmed investors · "DMind #1 on Hugging Face" claim has no link · funding amount undisclosed.

6. Key risks

  • Technical: audit's critical was a transaction-system error; agents placing real perp orders turn bugs into user losses. Audit scope = web platform, not CLI/Autopilot.
  • Incentives: paid points attract farmers; generous conversion = dump, stingy = angry users.
  • Governance: no token/DAO/treasury design published.
  • Legal: AI trade execution + a token tied to paid points may raise advice/custody/securities questions (US-HQ team).
  • Venue concentration: perps depend on Hyperliquid + Lighter.

7. Disclosure I hold a small Sparks balance from earlier use (no active paid plan). No token (none exists), no investment, not compensated. I run my own automated trading bots on Hyperliquid, which overlaps with Minara's Autopilot use case.

Primary sources

Questions and pushback welcome, especially on the Sparks-as-presale point.

Attached source files