Adding a new detector

Slither's plugin architecture lets you integrate new detectors that run from the command line.

Detector Skeleton

The skeleton for a detector is:

from slither.detectors.abstract_detector import AbstractDetector, DetectorClassification


class Skeleton(AbstractDetector):
    """
    Documentation
    """

    ARGUMENT = 'mydetector' # slither will launch the detector with slither.py --detect mydetector
    HELP = 'Help printed by slither'
    IMPACT = DetectorClassification.HIGH
    CONFIDENCE = DetectorClassification.HIGH

    WIKI = ''

    WIKI_TITLE = ''
    WIKI_DESCRIPTION = ''
    WIKI_EXPLOIT_SCENARIO = ''
    WIKI_RECOMMENDATION = ''

    def _detect(self):
        info = ['This is an example']
        res = self.generate_result(info)

        return [res]

• ARGUMENT lets you run the detector from the command line
• HELP is the information printed from the command line
• IMPACT indicates the impact of the issue. Allowed values are:
  • DetectorClassification.INFORMATIONAL: printed in green
  • DetectorClassification.LOW: printed in green
  • DetectorClassification.MEDIUM: printed in yellow
  • DetectorClassification.HIGH: printed in red
• CONFIDENCE indicates your confidence in the analysis. Allowed values are:
  • DetectorClassification.LOW
  • DetectorClassification.MEDIUM
  • DetectorClassification.HIGH
• WIKI constants are used to generate automatically the documentation.

_detect() needs to return a list of findings. A finding is an element generated with self.generate_result(info), where info is a list of text or contract's object (contract, function, node, ...)

An AbstractDetector object has the slither attribute, which returns the current Slither object.

Integration

You can integrate your detector into Slither by:

• Adding it in slither/detectors/all_detectors.py
• or, by creating a plugin package (see the skeleton example).

Test the detector

If you want to add your detector to trailofbits/slither, create a unit-test in tests and update scripts/travis_test_4.sh to run the unit-test automatically.

deepdiff is needed to run travis_test.sh:

pip install deepdiff

Example

backdoor.py will detect any function with backdoor in its name.