How it works

1. User Registration & Verification

  1. Users register on the platform via Celo’s Self protocol.

  2. The platform generates a QR code, which the user scans to initiate verification.

  3. An API call is sent to Self, generating a zero-knowledge proof containing a unique nullifier for the user.

  4. The nullifier is mapped to the user’s wallet address and stored in the smart contract.

  5. Users are now verified contributors, eligible to create or solve issues on Creo.

Key benefits:

  • Prevents fake accounts and Sybil attacks

  • Ensures every contribution is tied to a unique, verified human

2. Creating Issues

  1. A verified user submits a GitHub issue URL along with a bounty and metadata (difficulty, description, deadlines).

  2. A small portion of the bounty is automatically sent to the ASI agent for analysis.

  3. The issue is stored in the Creo smart contract, including deadlines, difficulty, and AI scoring parameters.

  4. The issue becomes visible on the platform, ready for contributors to take.

Contract Logic Highlights:

  • AI service fee is automatically collected

  • Deadlines are determined based on difficulty (easy, medium, hard)

  • Minimum completion percentage for stake return is defined

3. Taking & Staking an Issue

  1. A verified contributor chooses an issue to solve.

  2. Contributor stakes 5-20% of the bounty, ensuring commitment and reducing Sybil/DoS risk.

  3. The smart contract records:

    • Assigned contributor

    • Stake amount

    • Deadline

    • Contributor history

Key protections:

  • Creator cannot assign the issue to themselves

  • Contributors cannot take the same issue multiple times

  • Deadlines prevent stalling or abuse

4. AI-Powered Grading & Code Review

  1. ASI agents analyze the issue and provide a confidence score for the work done.

  2. Multi-agent review ensures robust, unbiased evaluation of contributions.

  3. The score is stored on-chain to influence final reward distribution.

5. Completing an Issue

  1. Once work is done, the creator approves completion through the platform.

  2. The contributor’s stake is returned, along with the bounty (adjusted for AI scores if needed).

  3. If the contributor fails to meet the minimum completion percentage, a fraction of their stake may be forfeited and added to the bounty.

6. Claiming Expired or Disputed Issues

  1. If the deadline expires and the issue is not completed:

    • The contributor can claim partial reward based on completion percentage

    • The stake is adjusted according to completion rules

  2. Platform tracks all expired or disputed issues automatically, without manual intervention.

7. Increasing Bounty or Extending Deadlines

  • Creators can increase the bounty or extend deadlines if needed, by interacting with the smart contract.

  • All changes are transparent and enforceable by contract rules.

8. On-Chain Transparency

  • All issues, stakes, nullifiers, and rewards are stored on-chain.

  • Contributors can always view:

    • Their assigned issues

    • Previous contributors for an issue

    • Remaining bounty or stakes

  • Smart contracts ensure automatic enforcement of rules, eliminating trust dependency.

9. Off-Chain Agents & Sybil Resistance

  • Contributors can deploy off-chain code solver agents to work on issues.

  • Agents interact with the platform through staking and deadlines.

  • Verification via nullifiers ensures only unique human contributions are counted.

Result: A fully trustless, fair, and secure platform for bounty-based contributions with verifiable human and AI oversight.

PreviousCore ConceptsNextArchitecture

Last updated