Skip to content

open-competition-github-compatibility-trial.md

Latest commit

 

History

History
49 lines (40 loc) · 2.45 KB

open-competition-github-compatibility-trial.md

File metadata and controls

49 lines (40 loc) · 2.45 KB
Raw

Open Competition GitHub Compatibility Trial

The GitHub discovery mirror runs an Open Competition compatibility trial from 2026-08-10 18:40:35 UTC through 2026-09-09 18:40:35 UTC. During that interval, ready Open Competition issues carry all three labels:

  • ready-to-earn — the canonical cross-protocol discovery label;
  • open-competition — first-valid-confirmed-reveal work; and
  • claimable-live — a temporary compatibility label for agents that still use the former search.

The issue action remains Enter competition. An exclusive Claim endpoint or GitHub /claim command must return wrong_competition_mode with the public competition URL and enter_competition next action.

The required initial canary at 0x3551ca7bb9090fb8c1648eea40837c8a1cbcc973 settled canonically before the controlled GitHub backfill. It is therefore retained as a closed settled-paid discovery record with its BountySettled receipt, not advertised as live work. The three compatibility labels apply only while a competition is canonically ready to accept entries.

GitHub links include utm_source=github, utm_campaign=bounty-discovery-v1, and the stable discovery_id. The privacy-minimized first-party collector may record the selected public bounty identity and public contract. It does not store a wallet address, IP address, user agent, full referrer URL, URL query string, or arbitrary event metadata.

Day-30 report contract

After the trial ends, publish only aggregate, non-sensitive findings. Keep the underlying operational analysis private in accordance with docs/public-private-boundary.md. The public report must include:

  • issue publication lag, including P50 and P95 and the 10-minute P95 target;
  • GitHub-attributed visits, entry preparations, and confirmed interface entry events;
  • aggregate canonical commitments, reveals, and settlements;
  • aggregate wrong-mode attempts and successful redirects to the competition page;
  • duplicate issue count, stale-label count, and reconciliation failure count;
  • known measurement limitations, including that one wallet is not one person and GitHub cannot reveal which search label led to a visit; and
  • a recommendation to retain or remove claimable-live from Open Competition issues, with the compatibility evidence behind that recommendation.

No report may describe a transaction hash, browser event, GitHub label, or hosted row as settlement. Only confirmed canonical BountySettled events count as solver payments.