-
Notifications
You must be signed in to change notification settings - Fork 982
Adding a new detector
Slither's plugin architecture lets you integrate new detectors that run from the command line.
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!'
finding = self.generate_json_result(info)
return [finding]
-
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 a dictionary that can be generated with self.generate_json_result(info)
, where info
is the text that is going to be printed.
The following helper allows to add information to the finding, that can be used by a continuous integration system or an IDE to locate the bug:
self.add_variable_to_json(variable)
self.add_variables_to_json(variables)
self.add_contract_to_json(contract)
self.add_function_to_json(function)
self.add_functions_to_json(functions)
self.add_nodes_to_json(nodes)
An AbstractDetector
object has the slither
attribute, which returns the current Slither
object.
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).
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
backdoor.py will detect any function with backdoor
in its name.