FromAI Agent Trust Layerby ZiggyPayoutby Pvium

Know which pull requests deserve maintainer trust.

AI agent trust layer for contributor reputation, PR risk analysis, and reward workflows.

Ziggy is a GitHub-native AI agent for contribution intake. It scores contributor reputation, flags risky pull requests, records maintainer decisions, and gives teams a clean audit trail before work is accepted or rewarded. Pvium powers payout when the project wants to pay genuine contributors anywhere in the world.

Coming soonFilter contributors by verified identity, powered by Pvium KYC.

0

Trusted Contribution Intake

Open-source projects use Ziggy to decide when a reward should be paid. Internal teams can use the same signal to protect review time: who is sending the change, what risk the diff introduces, and how that decision should be recorded.

Contributor reputation

Score incoming contributors with GitHub history, merged work, account age, contribution patterns, and prior payout outcomes.

PR risk analysis

Surface spam, bot-like behavior, low-quality submissions, sensitive-path changes, and likely slop before reviewers spend time.

Maintainer decision support

Post a clear GitHub-native signal on the pull request while keeping the final accept, reject, and payout call with humans.

Auditable outcomes

Keep an append-only trail of inputs, scores, maintainer decisions, and reward outcomes so every contribution decision can be replayed.

1

Add Ziggy To Your GitHub Repo

  1. Open the Ziggy GitHub App page.
  2. Choose the GitHub organization or account that owns your project.
  3. Select the repositories where maintainers should be able to pay rewards.
  4. Finish the installation and start adding reward labels to issues.
2

Add Rewards

Add a Ziggy reward label to any issue you want to pay for. This deployment uses the label prefix ziggytest:.

ziggytest:10USDCziggytest:25USDCziggytest:100USDC
3

Review Relevant PRs

The dashboard is where Ziggy turns noisy GitHub activity into a maintainer queue. Instead of opening every PR, you see the submissions that deserve review first.

  1. Open the maintainer dashboard to see rewarded issues grouped with their related PRs.
  2. Ziggy filters out late, low-signal, spammy, copied, or risky submissions by default.
  3. Review the PRs that surface with the strongest contributor and code-risk signals.
  4. Open the details when a submission needs closer inspection, then keep the final merge and payout decision with the maintainer.
4

Pay Contributors

  1. Approve the relevant PR after review and merge it into the configured reward branch.
  2. Ziggy confirms the PR closes the rewarded issue and passes the maintainer's payout threshold.
  3. The contributor connects their GitHub account to Pvium if they have not already done so.
  4. Pvium creates the payment flow, settles the reward, and Ziggy marks the reward paid.
5

Reward Rules

  • The pull request must be merged into a configured reward branch, usually main or master.
  • The pull request must close the rewarded issue with a GitHub closing reference.
  • The contributor must connect the same GitHub account that authored the merged PR.
  • After a contributor links Pvium once, future rewards can create payment links without another invite.
  • Ziggy scores risk signals, but maintainers stay in the middle for payment decisions.