Sybil resistance for the agentic web.
Powered by BringID. Native to EIP-8004.

One attacker can spin up a thousand agents, drain your faucet, flood your API, and game your reputation system. BringID enforces one credential per agent — on-chain, private, unforgeable.

Without verification
𝗐
attacker
b
b
b
b
b
b
SERVICE
overloaded
▲ 6 bots flooding requests
𝗐
human
b
✕
denied
▼ legitimate agent starved of resources

One attacker, six bots. The service can't tell who's real. Legitimate agents get crowded out.

With BringID verification
𝗐
attacker
1:1
b
b
b
SERVICE
1 share
▲ 1 credential = 1 agent. 5 bots rejected.
𝗐
human
1:1
b
SERVICE
1 share
▼ fair access. each human gets equal share.

One credential per agent. Attackers can't multiply. Every human-backed agent gets fair access.

@a16zcryptoAI needs blockchains — the internet has no native way to separate humans from machines↗@galnagliAgent networks are vulnerable to sybil attacks — we need proof there’s a human behind each agent↗
For services & protocols
Stop sybil attacks

Check if an agent has a unique, verified owner before granting access. One contract read. If the score is zero, the agent is unverified — treat accordingly.

// check agent's sybil resistance score const { totalScore } = await validator.getAgentScore( BigInt(agentId) ); if (totalScore > 0) { // unique verified owner ✔ }
npm @bringid/validator8004github BringID/bringid-validator-8004spec Integration docs
For agent owners
Verify your agent

Bind your identity to your EIP-8004 agent. Sybil-protected services only accept verified agents — one credential, one agent, no exceptions.

Read https://8004.bringid.org/skill.md and follow the instructions to verify your agent's humanity.
47 agents verified
What sybil resistance unlocks

Testnet faucets

Gate faucet access on verification. One person can't drain the pool with a thousand agents.

Spam defense

Filter inbound agent requests. Reject unverified agents before they hit your service.

API access tiers

Tier rate limits by verification level. Verified agents get priority access.

Airdrop eligibility

Token distributions require a unique verified owner. No more farming with duplicate agents.

How it works
1
Agent owner verifies credentials through BringID — X, GitHub, Farcaster, Binance KYC, zkPassport. Each credential binds to one agent only.
2
Score and nullifier recorded on-chain in the EIP-8004 Validation Registry. Credentials can’t be reused for another agent.
3
Services query the score with one contract read — no API keys, no trust assumptions. Zero score = unverified.
The details
Multi-credential
X, GitHub, Farcaster, Binance KYC, zkPassport — each adds to the score. More credentials = harder to fake.
1:1 binding
One credential = one agent. Proofs cryptographically bound to agentId. Can’t be reused, split, or frontrun.
Private
MPC-TLS proofs. No personal data on-chain. Only scores and nullifiers. The service knows you’re unique — not who you are.
Native EIP-8004
Lives in the Validation Registry. No sidecars, no extra infra. Query it like any other validation.