← back to dashboard · single-rule playground →
A fact pattern walks through three stages: pick a claim type, fill in the facts each applicable rule needs, and the simulator runs every rule against the same fact pattern and aggregates the result. Each rule is encoded in rules/ as a Catala module; the inputs are auto-generated from catala json-schema; the predicates run server-side via catala interpret -F json. The dispositive question — "if filed today at this tribunal, what does the rule say?" — is computed mechanically.
Six claim types coded across the corpus. Selection drives which rule modules light up.
Required facts come from the rule modules' input schemas. Defaults are seeded; edit anything.
Run the rules to see the aggregated verdict.