Sybil-Resistant Token Distribution¶

The Airdrop Problem¶

Most crypto protocols distribute tokens to the wrong people.

Not intentionally — but because the tools used to qualify recipients (wallet age thresholds, on-chain activity minimums, snapshot balances) are trivially gamed by anyone who has done it before. A professional airdrop farmer can spin up hundreds of wallets, meet every eligibility criterion, and dump every token the moment trading opens. The protocol wastes its treasury. The community is diluted. The price chart tells the story.

The numbers are stark: Uniswap's UNI airdrop sent ~$6M to Sybil clusters. Optimism's second airdrop estimated 17% of recipients were Sybil addresses. Arbitrum identified and excluded over 150,000 Sybil wallets — but only after significant manual investigation.

Manual Sybil analysis doesn't scale. Behavioural screening does.

Why Standard Eligibility Criteria Fail¶

Criterion	How it's gamed
Wallet age > 6 months	Farm wallets are created years in advance
Minimum transaction count	Scripted transactions cost pennies
On-chain activity on target protocol	Bots interact with every major protocol
Snapshot token balance	Balances are borrowed for the snapshot window
Cross-chain activity	Multi-chain scripts exist for every major bridge

Every rule-based eligibility filter creates a farming target. Once the criteria are known, the farms adapt within hours.

The only signal that can't be cheaply replicated: years of genuine, diverse, capital-backed on-chain participation across multiple protocols. This is exactly what Wallet Rank measures.

ChainAware's Approach: Behavioural Qualification¶

Instead of checking whether a wallet meets criteria, ChainAware assesses whether a wallet behaves like a genuine participant.

Every wallet in your distribution list is scored across ten behavioural parameters:

Experience — years of active on-chain engagement across multiple protocols
Fraud probability — AI model trained on 14M+ wallets, 98% accuracy
AML status — mixer interactions, sanctioned addresses, darknet exposure
Risk profile — coherent investment behaviour vs. scripted farm patterns
Protocol diversity — genuine DeFi activity vs. narrow scripted interactions
Wallet age — cannot be accelerated; a 3-year-old wallet required 3 years
Capital deployment — real financial skin-in-the-game, not dust transactions

Wallets that score poorly on behavioural parameters are excluded, regardless of whether they technically meet eligibility criteria. Wallets that score well receive merit-weighted allocations.

Use Cases¶

Airdrop Campaigns¶

Screen every recipient wallet before distribution. Receive three lists:

✅ Eligible — genuine participants, ranked by reputation for merit-based allocation
⚠️ Flagged — borderline wallets recommended for manual review
🚫 Disqualified — bots, fraud wallets, fresh Sybil clusters, AML-flagged addresses

A typical distribution to 1,000 wallets surfaces 15–25% disqualification — saving those tokens for genuine community members or future distributions.

IDO Whitelist Qualification¶

Before opening an IDO to registered participants, batch-screen the entire whitelist. Apply minimum Wallet Rank thresholds to ensure allocations go to experienced, long-term participants rather than opportunistic flippers.

Protocols that implement behavioural qualification report significantly lower day-one sell pressure — because the recipients are holders, not traders.

Fair Launch Allocations¶

For protocols prioritising community ownership over speculation, use reputation-weighted allocations: wallets with higher Wallet Rank scores receive proportionally larger allocations, ensuring the most engaged and legitimate participants get the largest stake.

Ongoing Community Qualification¶

Beyond the initial distribution, use behavioural screening to qualify participants for:
- Governance rights (minimum Wallet Rank threshold)
- Liquidity incentives (exclude farm wallets from reward programmes)
- Community roles (DAO membership, working groups)
- Staking tiers (higher tiers require higher reputation scores)

The `chainaware-airdrop-screener` Agent¶

The dedicated screening agent handles the full workflow — from raw address list to ranked eligible recipients with suggested token allocations.

Screen these 500 wallets for our airdrop — filter bots and fraud.
Budget: $200,000 in tokens | Network: Ethereum
[list of addresses]

Fair airdrop for our community round — merit-weighted allocations:
[1,000 addresses] on BNB Chain | Budget: $500,000

Sybil filter for our governance airdrop — minimum reputation score 2,000:
[addresses] on Ethereum

Example output:

AIRDROP SCREENING REPORT
Input: 500 wallets | Budget: $200,000 | Network: Ethereum

✅ Eligible:     347 (69%)
⚠️ Flagged:       48 (10%) — manual review
🚫 Disqualified: 105 (21%)
  → High fraud risk:     41
  → New/fresh wallets:   38
  → Bot patterns:        18
  → AML flags:            8

MERIT-WEIGHTED ALLOCATIONS
  Top tier   (rep >2,500):  62 wallets × $1,200 = $74,400
  Mid tier   (rep 1,000–2,499): 143 wallets × $450 = $64,350
  Base tier  (rep <1,000):  142 wallets × $150 = $21,300

Tokens saved from Sybil exclusion: ~$63,000

Combining With Token Launch Auditing¶

Sybil-resistant distribution is one half of a fair launch. The other half is verifying that the token itself — its contract and deployer — is trustworthy before asking the community to receive it.

The chainaware-token-launch-auditor runs both checks in a single call: contract rug pull risk + deployer wallet behavioural screening + a Launch Safety Score.

Pre-launch audit before we distribute:
Token: 0xContract... Deployer: 0xDev... Network: Ethereum

A APPROVED verdict confirms both that the distribution is going to genuine participants and that the token they are receiving is not itself a trap.

Integration¶

Via Prediction MCP¶

Batch screening integrates naturally into AI agent pipelines. The airdrop-screener and compliance-screener agents can process thousands of wallets in a single session, rank them by reputation, and output structured allocation tables.

claude mcp add --transport sse chainaware-behavioral-prediction \
  https://prediction.mcp.chainaware.ai/sse --header "X-API-Key: YOUR_KEY"

Via REST API¶

For automated pipelines processing registration data at scale:

POST /v1/fraud/batch
{ "addresses": ["0x...", "0x...", ...], "chain": "ethereum" }

Returns fraud scores, AML flags, and behavioural profiles for every address in the batch.