Reads columbus-5 directly from a public RPC node in your browser — no third-party dashboards. It pins the block at each UTC midnight, queries total uluna supply (net burns) and the community pool at that height, and derives taxable volume from the 1.5% tax split live since Aug 2, 2026 (prop #12223: 1.2% burn / 0.15% community pool / 0.15% oracle pool).
The green line is the page's benchmark: the typical daily taxable volume before the 1.5% tax — currently —. It is measured, not estimated: for each of the fourteen full weeks from Apr 19 to Jul 25, 2026, the actual tax burn is read from the blockchain and converted to volume at the 0.5%-era rate. Two of those weeks are dropped because they contain large one-off burns that were not tax (leaving twelve), weeks after voting on prop #12223 began are not used (activity around the vote was not typical), and the median — the middle week — is taken rather than the average, so a single unusual week cannot pull the benchmark up or down.
How to read it: bars above the line mean more on-chain volume than was typical before the rate change; bars below mean less. Some drop is expected at a higher rate — the deciding level is one-third of the line (−67%). Because the rate tripled, daily volume above that level means the 1.5% tax is still collecting and burning more than 0.5% would have; volume below it for a sustained period means the opposite. The verdict panel just below tracks exactly this.
Every embedded boundary supply was re-read from the chain on 2026-08-06 and cross-checked on a second independent node (publicnode); one boundary (Aug 2) was corrected by 63,220 LUNC in that audit. Method: day boundaries are found by walking block timestamps (~6s blocks) to within a minute of 00:00 UTC, then abci_query reads
cosmos.bank.v1beta1.Query/SupplyOf(uluna) and cosmos.distribution.v1beta1.Query/CommunityPool at each historical height.
This tool measures tax-driven on-chain burning only: known non-tax burns (e.g. Binance's monthly manual burn sends) are subtracted from
the supply change before anything is displayed, so they never appear in any chart, total, or implied-volume figure.
Taxable volume is estimated two ways: net burn ÷ burn share (conservative on quiet days, inflated by unknown one-off burns) and
CP inflow ÷ CP share (inflated by non-tax pool income). Tax collected = burn-implied volume × 1.5%. The impact view rate-normalizes both
eras (0.4%/0.05% shares before Aug 2, 1.2%/0.15% after) and compares medians of full days only, so one-off burns and the partial day don't skew
the baseline; the two estimators agree closely pre-1.5%-tax, which validates the method — after the 1.5% tax the CP-implied figure runs higher, so treat
burn-implied as the primary series. The weekly chart uses Sunday 00:00 UTC boundaries (the 1.5% tax went live on a Sunday, so eras align);
weekly tax = (weekly burn − known one-off burns) × 1.25, since 80% of the tax is burned in both eras. Public nodes prune old state; the deepest
found (stakely.io, the default here) reached ~mid-April 2026 when checked on Aug 6; the 0.5%-era weeks back to Apr 19 were read from it then and are embedded here (state below that was already pruned). The weeks of Apr 26 (~2.2B) and May 31 (~2.5B) are
deliberately omitted: they contained unidentifiable one-off burns that would misstate tax revenue (the week of May 3 also runs hot, ~1.3B, but is kept — the median is robust to it). Refreshing merges new weeks with this
embedded history rather than replacing it; each new 1.5% week accumulates automatically, and an archive node in the RPC box recovers deeper history. USD
figures use the editable price box; the last day/week shown is partial (up to the latest block). The daily range is automatic: it starts at the
Jul 28 baseline and extends itself one day at a time — refreshes fetch only days not already held, merging with the embedded history (refresh at
least every couple of months so days don't fall out of the node's retention window before being captured). Once the range exceeds 21 days the
daily charts show the latest 21 while totals, medians, and the table keep the full range.
If you find this tracker useful, a pizza is always welcome — LUNC, USTC or CL8Y:
terra1vuwsrt5rj405tq4kqx3knhzrxjcqtk9uxmt79n