Problem
The signature is final
Self-custody removes institutional reversals. A malicious approval, poisoned address, or drainer call becomes final the moment the user signs.
Whitepaper
A security layer that pays for itself per call.
A web-native reading of the Flick whitepaper: why pre-signature security must be local-first, machine-payable, and callable by both humans and autonomous agents.
Thesis
The market does not need another scanner. It needs security in the signing path.
Window
The only useful moment is before release
Simulation, calldata decoding, counterparty checks, and allowance review must happen inside the wallet prompt window, before the signature leaves the user.
Gap
Security exists, but not in the default path
Scanners and audit tools are fragmented, account-bound, and built for professionals. Flick turns them into one automatic local agent.
Agent
Flick is three functions under one local policy and one bounded budget.
Core utility
Inspect. Audit. Watch.
The dApp makes every on-chain action safer by default and pays for the intelligence required through a per-call budget.
Guard
Simulates the candidate transaction, decodes calldata, checks approvals, evaluates counterparty risk, and returns allow, warn, or block.
Audit
Buys an assessment for a specific contract, token, or protocol when the user or another agent actually needs the answer.
Watch
Re-evaluates standing exposure such as open approvals, positions, and protocol dependencies as new information appears.
Architecture
The complete financial graph stays local; only narrow questions leave the device.
Wallet / agent intent
Guard intercept
Policy + budget
Verdict
Simulation endpoint
Threat graph
Audit / inference
x402
The payment loop is the mechanism that removes accounts, API keys, and subscription friction.
1
Agent asks for one unit of work: a simulation, audit, threat lookup, or model inference.
2
Provider responds with 402 Payment Required, price, accepted asset, and destination.
3
Local policy checks balance and budget before any payment is authorized.
4
Agent signs exactly the required payment authorization and retries the request.
5
Provider returns the result, and the local spend accumulator updates.
Economics
Protection is rational when the per-call cost is lower than expected avoided loss.
CALCULATION
unprotected = qL
protected = (1-d)qL + C
C < dqL
The important design point is structural: cost is incurred per action, so users are not paying during long periods with no active risk.
Users pay only when risk appears
Cost tracks protective activity instead of a fixed subscription paid during quiet periods.
Providers earn per useful answer
Simulators, auditors, inference providers, and threat-data maintainers monetize the exact unit consumed.
Agents buy verdicts before signing
Trading, treasury, and assistant agents can call Flick as a machine-payable security primitive.
Token
$FLICK sits inside the security action as the settlement and metering instrument.
Settlement
$FLICK meters Guard, Audit, Watch, and verdict-as-a-service calls.
Distribution
95% liquidity and 5% marketing; no founder allocation, investor tranche, or discretionary treasury.
Demand
Utility demand follows the amount of intelligence consumed or verdicts sold by the network.
Sustainability
Endpoint operators are funded by per-verdict revenue; shared work can be funded from protocol fees.
Threat model
The security claim is narrow because defensible security must state its limits.
Designed for transaction-time deception: drainers, malicious approvals, poisoned addresses, misleading calldata, and compromised dependencies.
Not a guarantee. Novel exploits can pass when no consulted intelligence source detects them.
Local-first protects against third-party aggregation, but device compromise is outside the claim.
The agent gates a signature; it does not seize one. User override remains part of self-custody.
Roadmap
Each stage is useful on its own and compounds into an open provider market.
Stage I
Guard
In-path pre-signature checks with x402-metered simulation and threat lookup.
Stage II
Audit and Watch
On-demand contract assessment and continuous exposure monitoring under the same budget.
Stage III
Verdict-as-a-service
External agents purchase priced safety verdicts before delegated signing.
Stage IV
Open provider market
Permissionless intelligence and verdict providers with quality signaling.