SourceCred’s core reusable mechanism is the contribution graph: contributions, people, and their relationships become nodes and edges over which Cred flows rather than a flat score table.
The project is especially useful because it cleanly separates three layers that many systems blur together: value legibility (Cred), reward policy (Grain), and operator configuration (plugins, weights, identity merges, and instance administration).
Community control over node and edge weights is the main governance surface. It determines what actions mint Cred and how much downstream credit supporting work receives.
Retroactivity is a major design choice. SourceCred expects scores to change when communities add new data, adjust weights, or reinterpret what mattered.
Grain is not a fixed tokenomic formula. The docs expose multiple payout policies — immediate, balanced, and recent — that can yield very different labor-market effects from the same Cred history.
The setup docs show that SourceCred is also an operating model: plugins, GitHub Actions, GitHub Pages, and admin tasks define the practical boundary of what labor the system can see and reward.
Specific source signals worth retaining
Introduction docs: strongest concise statement of the project thesis — making labor visible and rewardable.
Cred docs: best explanation of the contribution graph, seed node, alpha/radiation model, and the difference between node weights and edge weights.
Grain docs: clearest source on payout policy, opt-in rewards, transferability, and the separation between Cred and actual compensation.
Setup guide: important because it reveals the real deployment architecture and the plugin/data-source constraints on any SourceCred instance.
Repos: useful confirmation that SourceCred is implemented as a monorepo plus separate docs repo rather than only a conceptual scoring framework.
Why this entry belongs in the corpus
It strengthens the public-goods and contributor-reward cluster with a canonical example of graph-based retroactive labor accounting.
It provides a strong comparison class for Praise, Drips, and Open Source Observer because each system makes labor legible in a different way: peer recognition, dependency graphs, project registries, or contribution-flow graphs.
The reusable insight is that the decisive governance surface often lives upstream of payouts: in event ingestion, identity resolution, weighting rules, retroactivity, and the policy that converts social graph signals into money or influence.