I'm a founder building a bug bounty settlement protocol (ZK-PoE, Sepolia testnet, not mainnet). The idea: you run the exploit against an authorized target, a notary signs the HTTP response, and a Groth16 proof shows the secret was captured. The payout settles on-chain in the same transaction that reveals your key. No triage queue.
Be blunt with me, since this is the community I'd lose if I'm wrong:
What would make you refuse to use this over HackerOne/Bugcrowd?
Is writing a request-spec JSON (method, path, headers, body) hard, or is the notarization step the real friction?
Would you trust an automated severity grader if the payout is bounded by a published rate table?
Disclosure: I built this, so weigh my answers accordingly. Happy to be told it's a bad idea.