What it measures: How consistently the contributor has been active over the analysis period.
Why it matters: Consistent activity over time indicates genuine engagement with open source. Burst activity patterns (e.g., many contributions in a single day) can be a sign of spam or AI-generated contributions.
Consistency Score = Months with Activity / Total Months in Analysis Window
Activity Definition:
- A month is considered "active" if the contributor created at least one PR during that month
- Only months within the analysis window are considered
Data Sources:
- GitHub GraphQL API:
user.pullRequests.nodes[].createdAt - Grouped by month to count unique active months
| Input | Default | Description |
|---|---|---|
threshold-activity-consistency |
0 |
Minimum consistency score (0-1) |
analysis-window |
12 |
Months of history to analyze |
For a 12-month analysis window:
- 12 months with activity: 100% consistency (1.0)
- 6 months with activity: 50% consistency (0.5)
- 1 month with activity: 8.3% consistency (0.083)
- New accounts with less history than the analysis window are evaluated on available months
- If account is newer than analysis window, only counts months since account creation
- Months with only closed (not merged) PRs still count as active
- Contribute regularly over time rather than in bursts
- Set aside time each month for open source contributions
- Start with small, manageable contributions you can sustain