Signal-ID coverage checker
Expose duplicate, missing, unexpected and orphan labels before interpreting declared evidence subsets.
Evidence methods, not a provider certification. Authored examples and user-entered check states are not independently verified trading results.
Exact membership before fractions
Supply a unique nonempty original roster and record rows. The checker compares exact case-sensitive IDs, reporting missing, duplicated and unexpected labels. It also checks declared version parents for absent links, different originals and cycles. Matching total rows cannot hide an ID discrepancy. Original, row and amendment counts are kept separate.
State labels are not checked proofs
Each row contains commitment and proof-file booleans, a declared proof state, an execution-availability boolean and a terminal label. The schema rejects inconsistent proof-file states and unsupported labels. Declared_pass requires a file and commitment label, but those booleans authenticate nothing. The tool does not run a Bitcoin verifier, inspect a broker statement or reproduce a trading return.
Exact fractions remain scoped
Only after identity and parent graph reconciliation does the export contain numerator/denominator pairs for commitment, proof-file, declared successful-check, execution-availability and joined subsets. A closed-only winner pair includes flats in its denominator. Unresolved identities retire ratios instead of shrinking the original population. Every export retains source_verified false and real-world validation fields null.
Bounds and private drafts
Each input array is limited to 1,000 entries. IDs use one to eighty ASCII letters, digits, periods, underscores, colons or hyphens, beginning with a letter or digit. Versions share that teaching syntax. The form accepts JSON arrays only and exports no automatic server submission. A reload restores the labelled example, not your private previous draft.