Paragon

Compliance, per cluster

Each criterion below compares a measured value against the threshold the program published for that cluster and epoch. The dashed line is the limit; the bar is the measurement. Nothing here is hardcoded: change the epoch and the limits move.

Program grade history

The grade the program assigned for each epoch it assessed. Bonus met the higher bar that earns additional stake; Baseline met the requirement without it; Nonemeans the epoch’s requirements were not met. An epoch with no grade at all was not assessed: it counts as neither pass nor fail, and renders as an empty track. Select a tile to read that epoch’s detail, including the failure reason where the program published one.

• Bonus: passed, full-height bar
• Baseline: passed, 60% bar
• Not met (None): failed, low bar
• Not graded: not assessed, empty track

“Graded span” is the first and last epoch the program published a grade for. For a validator no longer participating, the last graded epoch is strong evidence of when that ended, but the program states no exit date, so none is claimed here. Dates prefixed ~ are interpolated between known epoch boundaries.

Failure reasons are the program’s own machine-readable codes, collected by this site into a committed file; collection is still running, and the program published no reasons at all for most failures before roughly epoch 600. An unexplained failed epoch is still a failed epoch.

Build time

Median time the leader took to build a block on mainnet-beta, from IBRL, not from the Solana Foundation: the program publishes no measure of block building at all, so this is one of two figures here that are not its own, alongside Trillium slot duration.

Vote credit history

The last 64 epochs on each cluster, as a share of the credits achievable that epoch. The dashed line is the program’s floor converted to that same measure, so the two are genuinely comparable. Each axis is zoomed to its own data, with its range stated beneath the plot.