dGrants primary sources — 2026-05-09
What I reviewed
- Official repository:
- Detailed design specification in the repo:
Key takeaways from primary sources
- dGrants is not mainly interesting for reusing quadratic funding. Its main contribution is architectural: it decomposes a grants platform into a permissionless registry, separate round operators, pluggable curation, public indexing, and forkable UI.
- The design spec is explicit that the goal is openness and non-intervention by the Gitcoin team. That makes dGrants useful as a corpus entry even if it is partly historical, because it surfaces where centralization was perceived to live in the earlier grants stack.
- The registry design is intentionally minimal. Onchain records hold only owner, recipient, and metadata hash, while richer project data lives in content-addressed storage. This cleanly separates immutable identity and payout endpoints from mutable project presentation.
- The spec repeatedly acknowledges that permissionless registries create spam and fraud pressure, which is why curation is pulled out into its own layer rather than hidden inside one operator’s backend.
- Grant lists are one of the strongest ideas in the document. They adapt the Uniswap token-list pattern to grants discovery, letting different actors publish trusted collections that users can mix and match.
- The proposed prediction-market curation layer is also worth retaining. Even if it looks operationally heavy, it makes a sharp point: grant discovery and legitimacy can be treated as a market for revealed information rather than only a moderation queue.
- Round creation is decentralized in principle, but final distribution power still sits with the matching-pool custodian. That is an important governance reminder for comparisons with later systems such as Allo or managed grants platforms.
- The indexer and UI sections show that legibility is part of the protocol surface. If users rely on a public GraphQL indexer and a browser app, whoever defines those data models still shapes what becomes visible and fundable.
Specific source signals worth retaining
- README: the repo is a monorepo with contracts and frontend, and explicitly invites contributors to help build a decentralized Gitcoin.
- Objectives section: permissionless, transparent, secure, affordable, composable.
- Architecture section: grant registry, rounds and matching, curation, indexer, web UI, user store.
- Registry section: immutable minimal onchain records plus IPFS-backed metadata and update-by-replacement rather than in-place edits.
- Curation section: grant lists, user-selectable views, and experimental prediction-market ranking.
- Indexer section: The Graph-based indexing, public GraphQL API, and the idea that derived public data should be reconstructible by anyone.
- Web UI section: SPA-like architecture intended to be forkable and even hostable from IPFS.
Why this entry belongs in the corpus
- dGrants belongs because it makes the usually hidden control surfaces of grants infrastructure unusually explicit.
- It is especially useful next to Gitcoin Grants Stack and Allo because it asks what happens when the registry, curation, indexing, and interface layers are all treated as separable public infrastructure instead of bundled into one grants product.
- The reusable insight is that public-goods funding systems do not stop at the matching mechanism; they also depend on who curates lists, who runs indexers, who defines metadata schemas, and who can start or settle rounds.