Skip to main content

Out-Of-Range Signals

Detector Type:Compute or Constrain

Summary and Usage​

The Out-Of-Range Signals (OORS) detector analyzes patterns of computation and constraint generation to estimate the expected ranges of intermediate values (e.g., if a value is part of a byte decomposition, it should lie within [0, 255]). These expected ranges are then compared against the actual ranges enforced by constraints. If the expected range is narrower than the enforced range, values may fall "out of range," leaving them underconstrained and potentially introducing bugs or vulnerabilities.

Usage​

Select "Out-Of-Range Signals" in the "Required Detector Selection" section of the ZK Vanguard Task Wizard.

Example and Explanation​

The following example circom file contains the implementation of the U16 component, which is designed to compute a 16-bit value given two 8-bit inputs.

out_of_range_bug.circom
pragma circom 2.1.8;

template U16() {
  signal input highByte;
  signal input lowByte;
  signal output shortVal;

  // Assert that `shortVal` is 16-bit
  component lt = LessThan(252);
  lt.in[0] <== shortVal;
  lt.in[1] <== 65536;
  lt.out === 1;

  shortVal <== lowByte + (256 * highByte);
}

component main = U16();

While shortVal is properly constrained to be a 16-bit value and to equal the composition of lowByte and highByte, neither lowByte nor highByte is constrained to lie within the 8-bit range. As a result, multiple assignments of lowByte and highByte can yield the same shortVal. For example:

  • lowByte = 0, highByte = 1, shortVal = 256
  • lowByte = 256, highByte = 0, shortVal = 256 Both satisfy the constraints, even though only the first reflects the intended design.

Limitations​

  • This detector uses LLZK’s intraprocedural range analysis to infer expected intervals. Any inaccuracies in this analysis (e.g., when constraints are applied indirectly in other parts of the circuit) may result in false positives.
  • This detector only recognizes a specific set of patterns to infer likely intervals. It will not detect out-of-range errors for computation patterns that are highly circuit- or application-specific.
  • The analysis currently assumes the circuit operates over the BN254 field. We plan to enable the prime field to be configurable in an upcoming AuditHub release.

How to Assess Severity​

If confirmed as a true positive, a finding from the OORS detector indicates a severe issue. Underconstrained values may allow multiple different proofs to yield the same output, undermining the integrity of the circuit.