Run
Falsification campaign — 27 candidate(s), runner=modal
0
ruled out
3
still standing
24
needs more
Nothing was promoted.
A run only produces evidence and verdicts — it never advances a candidate on its own. A human operator holds the terminal accept/promote decision.
What we tried to kill — 27/27 run
24 more candidates had no runnable proposal — nothing was tested (no verified inputs to dispatch). That refusal is the spend gate working, not a result.