def match(self):
"""Match observables against Yeti's intelligence repository.
Takes an array of observables, expands them and tries to match them against specific indicators or known observables.
To "expand" an observable means to enrich the query. For instance, if the arrays of observables contains the URL ``http://google.com``,
the "expanded" observable array will also include the hostname ``google.com``.
:<json [string] observables: An array of observables to be analyzed
:>json [Entity] entities: Related ``Entity`` objects
:>json [Observable] known: ``Observable`` objects that are already present in database
:>json [Indicator] matches: ``Indicators`` that matched observables
:>json Observable matches[].observable: The ``Observable`` object that matched the ``Indicator``
:>json string unknown: Array of observable strings that didn't match any ``Indicators`` and are unknown to Yeti
"""
params = request.json
observables = params.pop('observables', [])
fetch_neighbors = params.pop('fetch_neighbors', True)
add_unknown = bool(params.pop('add_unknown', False))
if add_unknown and current_user.has_permission('observable', 'write'):
for o in observables:
Observable.add_text(o)
data = match_observables(observables, save_matches=add_unknown and current_user.has_permission('observable', 'write'), fetch_neighbors=fetch_neighbors)
return render(data)
def match(self):
params = request.json
observables = params.pop('observables', [])
add_unknown = bool(params.pop('add_unknown', False))
if add_unknown:
for o in observables:
Observable.add_text(o)
data = match_observables(observables, save_matches=add_unknown)
return render(data)
def derive(observables):
"""Indicate that the module needs a specific attribute to work properly.
This function is only useful in abstract modules, in order to make sure
that modules that inherit from this class correctly defines needed class
attributes.
Args:
variables: a string or an array of strings containing the name of
needed class attributes.
Raises:
ModuleInitializationError: One of the needed attributes is not
correctly defined.
"""
new = []
observables = list(iterify(observables))
for i, observable in enumerate(observables):
try:
t = Observable.guess_type(observable)
temp = t(value=observable)
temp.clean()
observable = temp.value
observables[i] = observable
for a in analyzers.get(t, []):
new.extend([n for n in a.analyze_string(observable) if n and n not in observables])
except ObservableValidationError:
pass
if len(new) == 0:
return observables
else:
return observables + derive(new)
def analyze(self, dict):
observable = dict['title']
description = dict['description'].lower()
context = {}
context['description'] = "{} C2 server".format(description)
context['date_added'] = datetime.strptime(dict['pubDate'], "%d-%m-%Y")
context['source'] = self.name
try:
e = Observable.add_text(observable)
except ObservableValidationError as e:
logging.error(e)
return
e.add_context(context)
e.add_source("feed")
tags = ['malware', 'c2', description, 'crimeware']
if description == 'pony':
tags.extend(['stealer', 'dropper'])
elif description == 'athena':
tags.extend(['stealer', 'ddos'])
elif description in ['zeus', 'citadel']:
tags.extend(['banker'])
e.tag(tags)
def analyze(observable, results):
links = set()
params = {'query': observable.value}
data = PassiveTotalApi.get('/dns/passive', results.settings, params)
for record in data['results']:
first_seen = datetime.strptime(
record['firstSeen'], "%Y-%m-%d %H:%M:%S")
last_seen = datetime.strptime(
record['lastSeen'], "%Y-%m-%d %H:%M:%S")
new = Observable.add_text(record['resolve'])
if isinstance(observable, Hostname):
links.update(
observable.link_to(
new, "{} record".format(record['recordType']),
'PassiveTotal', first_seen, last_seen))
else:
links.update(
new.link_to(
observable, "{} record".format(record['recordType']),
'PassiveTotal', first_seen, last_seen))
return list(links)
def derive(strings):
values = set()
observables = set()
for string in iterify(strings):
if string:
try:
t = Observable.guess_type(string)
observable = t(value=string)
observable.normalize()
observables.add(observable)
values.add(observable.value)
except ObservableValidationError:
values.add(string)
new = []
for observable in observables:
for a in analyzers.get(observable.__class__, []):
new.extend([
n for n in a.analyze_string(observable.value)
if n and n not in values
])
if len(new) == 0:
return values, values
else:
_, extended = derive(new + list(values))
return values, extended
def post(self):
q = request.get_json(silent=True)
params = q.pop("params", {})
observables = []
for o in q["observables"]:
try:
obs = Observable.guess_type(o['value'])(value=o['value'])
obs.clean()
observables.append(obs.value)
# Save observables & eventual tags to database
if params.get('save_query', False):
obs = obs.save()
obs.tag(o.get("tags", []))
obs.add_source("query")
except ObservableValidationError:
continue
# match observables with known indicators
data = match_observables([o for o in observables])
# find related observables (eg. URLs for domain, etc.)
# related_observables = [obs.get_related() for obs in observables]
# data = self.match_observables(related_observable)
#
# we need to find a way to degrade the "confidence" in
# hits obtained from related observables
return render(data, "analysis.html")
def analyze_outdated(self):
# do outdated logic
fltr = Q(**{"last_analyses__{}__exists".format(self.name): False})
if self.EXPIRATION:
fltr |= Q(**{"last_analyses__{}__lte".format(self.name): datetime.now() - self.EXPIRATION})
fltr &= Q(**self.CUSTOM_FILTER) & Q(_cls__contains=self.ACTS_ON)
self.bulk(Observable.objects(fltr))
def execute(self):
self.export_file_handle = codecs.open(self.output_file, 'w+', "utf-8")
q = Q(tags__name__in=[t.name for t in self.include_tags]) & Q(tags__name__nin=[t.name for t in self.exclude_tags])
q &= Q(_cls__contains=self.acts_on)
output = self.template.render(Observable.objects(q))
self.write(output)
self.export_file_handle.close()
def each(url):
try:
host = ProcessUrl.analyze_string(url.value)[0]
h = Observable.guess_type(host).get_or_create(value=host)
h.add_source("analytics")
Link.connect(src=url, dst=h)
except ObservableValidationError:
logging.error("An error occurred when trying to add {} to the database".format(host))
def each(url):
try:
host = ProcessUrl.analyze_string(url.value)[0]
h = Observable.guess_type(host).get_or_create(value=host)
h.add_source("analytics")
url.active_link_to(h, "hostname", "ProcessUrl", clean_old=False)
return h
except ObservableValidationError:
logging.error("An error occurred when trying to add {} to the database".format(host))
def enrich(self):
return "ENRICH"
if request.method == "POST":
lines = request.form['bulk-text'].split('\n')
for l in lines:
obs = refang(l.split(',')[0])
tags = refang(l.split(',')[1:])
o = Observable.add_text(obs)
o.tag(tags)
return render_template('observable/query.html')
def match_observables(observables):
# Remove empty observables
observables = [observable for observable in observables if observable]
extended_query = set(observables) | set(derive(observables))
added_entities = set()
data = {"matches": [], "unknown": set(observables), "entities": [], "known": [], "neighbors": []}
for o in Observable.objects(value__in=list(extended_query)):
data['known'].append(o.info())
del_from_set(data['unknown'], o.value)
for link, node in (o.incoming()):
if isinstance(node, Observable):
if (link.src.value not in extended_query or link.dst.value not in extended_query) and node.tags:
data['neighbors'].append((link.info(), node.info()))
for o, i in Indicator.search(extended_query):
o = Observable.add_text(o)
match = i.info()
match.update({"observable": o.info(), "related": [], "suggested_tags": set()})
for nodes in i.neighbors().values():
for l, node in nodes:
# add node name and link description to indicator
node_data = {"entity": node.type, "name": node.name, "link_description": l.description or l.tag}
match["related"].append(node_data)
# uniquely add node information to related entitites
if node.name not in added_entities:
nodeinfo = node.info()
nodeinfo['type'] = node.type
data["entities"].append(nodeinfo)
added_entities.add(node.name)
o_tags = o.get_tags()
[match["suggested_tags"].add(tag) for tag in node.generate_tags() if tag not in o_tags]
data["matches"].append(match)
del_from_set(data["unknown"], o.value)
return data
def analyze_outdated(self):
class_filter = Q()
for acts_on in iterify(self.ACTS_ON):
class_filter |= Q(_cls="Observable.{}".format(acts_on))
# do outdated logic
fltr = Q(**{"last_analyses__{}__exists".format(self.name): False})
if self.EXPIRATION:
fltr |= Q(**{"last_analyses__{}__lte".format(self.name): datetime.utcnow() - self.EXPIRATION})
fltr &= self.CUSTOM_FILTER & class_filter
self.bulk(Observable.objects(fltr).no_cache())
def post(self, action=None):
if action == 'merge':
tags = request.json['merge']
merge_into = Tag.objects.get(name=request.json['merge_into'])
make_dict = request.json['make_dict']
merged = 0
for tag in tags:
Observable.change_all_tags(tags, merge_into.name)
oldtag = Tag.objects.get(name=tag)
merge_into.count += oldtag.count
merge_into.produces += [i for i in oldtag.produces if i not in merge_into.produces and i != merge_into]
merge_into.save()
oldtag.delete()
merged += 1
if make_dict:
merge_into.add_replaces(tags)
return render({"merged": merged, "into": merge_into.name})
def post(self, id=None):
if not id:
data = request.json
data['produces'] = [Tag.get_or_create(name=t.strip()) for t in request.json['produces'].split(',') if t.strip()]
data['replaces'] = request.json['replaces'].split(',')
return render(Tag(**data).save().info())
else:
try:
data = request.json
data['produces'] = [Tag.get_or_create(name=t.strip()) for t in request.json['produces'].split(',') if t.strip()]
data['replaces'] = request.json['replaces'].split(',')
t = Tag.objects.get(id=id)
t.update(**data)
Observable.change_all_tags(t.name, data['name'])
return render({"status": "ok"})
except TagValidationError as e:
restful_abort(400, error=str(e))
except Exception as e:
import traceback
traceback.print_exc()
restful_abort(400, error='Must specify name and produces parameters')
def execute(self):
q_include = Q()
for t in self.include_tags:
q_include |= Q(tags__match={"name": t.name, "fresh": True})
q_exclude = Q(tags__name__nin=[t.name for t in self.exclude_tags])
q = (
Q(tags__not__size=0, tags__match={"fresh": True})
& q_include
& q_exclude
& Q(_cls="Observable.{}".format(self.acts_on))
)
return self.template.render(self.filter_ignore_tags(Observable.objects(q).no_cache()), self.output_file)
def nodesearch(self, query):
result = []
query = re.compile("^{}".format(query), re.IGNORECASE)
observables = Observable.objects(value=query).limit(5)
entities = Entity.objects(name=query).limit(5)
for results in [observables, entities]:
for node in results:
result.append(node.to_mongo())
return render(result)
def derive(observables):
new = []
for observable in iterify(observables):
try:
t = Observable.guess_type(observable)
for a in analyzers.get(t, []):
new.extend([n for n in a.analyze_string(observable) if n and n not in observables])
except ObservableValidationError:
pass
if len(new) == 0:
return observables
else:
return derive(new + observables)
def analyze(self, item, pub_date): # pylint: disable=arguments-differ
s_re = '\[([^\]]*)] Type: (\w+) - IP: (\d{1,3}.\d{1,3}.\d{1,3}.\d{1,3})'
r = re.compile(s_re)
m = r.match(item['description'])
malware_family = ''
c2_IP = ''
if m:
malware_family = m.group(2)
c2_IP = m.group(3)
observable = item['title']
description = item['description'].lower()
context = {}
context['description'] = "{} C2 server".format(c2_IP)
context['date_added'] = pub_date
context['source'] = self.name
c2 = None
e = None
try:
e = Observable.add_text(observable)
if c2_IP:
c2 = Ip.get_or_create(value=c2_IP)
e.active_link_to(c2, "IP", self.name, clean_old=False)
except ObservableValidationError as e:
logging.error(e)
logging.error(description)
return
tags = ['malware', 'c2', malware_family.lower(), 'crimeware']
if malware_family == 'pony':
tags.extend(['stealer', 'dropper'])
elif malware_family == 'athena':
tags.extend(['stealer', 'ddos'])
elif malware_family in ['zeus', 'citadel', 'lokibot']:
tags.extend(['banker'])
if e:
e.add_context(context)
e.add_source(self.name)
e.tag(tags)
if c2:
c2.add_context(context)
c2.add_source(self.name)
c2.tag(tags)
def derive(observables):
if isinstance(observables, (str, unicode)):
observables = [observables]
new = []
for observable in observables:
t = Observable.guess_type(observable)
for a in analyzers.get(t, []):
new.extend([n for n in a.analyze_string(observable) if n and n not in observables])
if len(new) == 0:
return observables
else:
return derive(new + observables)
def analyze(self, line):
if not line or line[0].startswith("#"):
return
date, _type, family, hostname, url, status, registrar, ips, asns, countries = tuple(
line)
item_date = dateutil.parser.parse(date)
max_age = yeti_config.get('limits', 'max_age')
limit_date = datetime.now() - timedelta(days=max_age)
if item_date < limit_date:
return
tags = []
tags += TYPE_DICT[_type]
tags.append(family.lower())
context = {
"first_seen": date,
"status": status,
"registrar": registrar,
"countries": countries.split("|"),
"asns": asns.split("|"),
"source": self.name
}
try:
url = Url.get_or_create(value=url.rstrip())
url.add_context(context)
url.tag(tags)
hostname = Observable.add_text(hostname)
hostname.tag(tags + ['blocklist'])
for ip in ips.split("|"):
if ip != hostname and ip is not None and ip != '':
try:
i = Ip.get_or_create(value=ip)
i.active_link_to(hostname,
"First seen IP",
self.name,
clean_old=False)
except ObservableValidationError as e:
logging.error("Invalid Observable: {}".format(e))
except ObservableValidationError as e:
logging.error(e)
def analyze_outdated(self):
class_filter = Q()
for acts_on in iterify(self.ACTS_ON):
class_filter |= Q(_cls="Observable.{}".format(acts_on))
# do outdated logic
fltr = Q(**{"last_analyses__{}__exists".format(self.name): False})
if self.EXPIRATION:
fltr |= Q(
**{
"last_analyses__{}__lte".format(self.name):
datetime.utcnow() - self.EXPIRATION
})
fltr &= self.CUSTOM_FILTER & class_filter
self.bulk(Observable.objects(fltr).no_cache())
def rdata_lookup(observable, api_key):
links = set()
for record in DNSDBApi.lookup('rdata', observable, api_key):
new = Observable.add_text(record['rrname'])
new.add_source('analytics')
links.update(
new.link_to(observable,
source='DNSDB Passive DNS',
description='{} record'.format(record['rrtype']),
first_seen=record['first_seen'],
last_seen=record['last_seen']))
return list(links)
def rdata_lookup(observable, api_key):
links = set()
for record in DNSDBApi.lookup('rdata', observable, api_key):
new = Observable.add_text(record['rrname'])
new.add_source('analytics')
links.update(
new.link_to(
observable,
source='DNSDB Passive DNS',
description='{} record'.format(record['rrtype']),
first_seen=record['first_seen'],
last_seen=record['last_seen']))
return list(links)
def each(cls, hostname, rtype=None, results=[]):
generated = []
h = Hostname.get_or_create(value=hostname.value)
for rdata in results:
logging.info("{} resolved to {} ({} record)".format(h.value, rdata, rtype))
try:
e = Observable.add_text(rdata)
e.add_source("analytics")
generated.append(e)
l = Link.connect(h, e)
l.add_history(tag=rtype, description='{} record'.format(rtype))
except ObservableValidationError as e:
logging.error("{} is not a valid datatype".format(rdata))
h.analysis_done(cls.__name__)
return generated
def rrset_lookup(hostname, api_key):
links = set()
for record in DNSDBApi.lookup('rrset', hostname, api_key):
for observable in record['rdata']:
observable = Observable.add_text(observable)
observable.add_source('analytics')
links.update(
hostname.link_to(observable,
source='DNSDB Passive DNS',
description='{} record'.format(
record['rrtype']),
first_seen=record['first_seen'],
last_seen=record['last_seen']))
return list(links)
def each(cls, hostname, rtype=None, results=[]):
generated = []
h = Hostname.get_or_create(value=hostname.value)
for rdata in results:
logging.debug("{} resolved to {} ({} record)".format(h.value, rdata, rtype))
try:
e = Observable.add_text(rdata)
e.add_source("analytics")
generated.append(e)
except ObservableValidationError as e:
logging.error("{} is not a valid datatype".format(rdata))
h.active_link_to(generated, "{} record".format(rtype), "ResolveHostnames")
h.analysis_done(cls.__name__)
return generated
def rrset_lookup(hostname, api_key):
links = set()
for record in DNSDBApi.lookup('rrset', hostname, api_key):
for observable in record['rdata']:
observable = Observable.add_text(observable)
observable.add_source('analytics')
links.update(
hostname.link_to(
observable,
source='DNSDB Passive DNS',
description='{} record'.format(record['rrtype']),
first_seen=record['first_seen'],
last_seen=record['last_seen']))
return list(links)
def rdata_lookup(observable, api_key):
links = set()
for record in DNSDBApi.lookup("rdata", observable, api_key):
new = Observable.add_text(record["rrname"])
new.add_source("analytics")
links.update(
new.link_to(
observable,
source="DNSDB Passive DNS",
description="{} record".format(record["rrtype"]),
first_seen=record["first_seen"],
last_seen=record["last_seen"],
)
)
return list(links)
def analyze(self, line):
if not line or line[0].startswith("#"):
return
date, _type, family, hostname, url, status, registrar, ips, asns, countries = tuple(
line)
tags = []
tags += TYPE_DICT[_type]
tags.append(family.lower())
context = {
"first_seen": date,
"status": status,
"registrar": registrar,
"countries": countries.split("|"),
"asns": asns.split("|"),
"source": self.name
}
try:
url = Url.get_or_create(value=url.rstrip())
url.add_context(context)
url.tag(tags)
hostname = Observable.add_text(hostname)
hostname.tag(tags + ['blocklist'])
for ip in ips.split("|"):
if ip != hostname and ip is not None and ip != '':
try:
i = Ip.get_or_create(value=ip)
i.active_link_to(
hostname,
"First seen IP",
self.name,
clean_old=False)
except ObservableValidationError as e:
logging.error("Invalid Observable: {}".format(e))
except ObservableValidationError as e:
logging.error("Invalid line: {}\nLine: {}".format(e, line))
def each(cls, hostname, rtype=None, results=[]):
generated = []
h = Hostname.get_or_create(value=hostname.value)
for rdata in results:
logging.debug("{} resolved to {} ({} record)".format(
h.value, rdata, rtype))
try:
e = Observable.add_text(rdata)
e.add_source("analytics")
generated.append(e)
except ObservableValidationError as e:
logging.error("{} is not a valid datatype".format(rdata))
h.active_link_to(generated, "{} record".format(rtype),
"ResolveHostnames")
h.analysis_done(cls.__name__)
return generated
def rrset_lookup(hostname, api_key):
links = set()
for record in DNSDBApi.lookup("rrset", hostname, api_key):
for observable in record["rdata"]:
observable = Observable.add_text(observable)
observable.add_source("analytics")
links.update(
hostname.link_to(
observable,
source="DNSDB Passive DNS",
description="{} record".format(record["rrtype"]),
first_seen=record["first_seen"],
last_seen=record["last_seen"],
)
)
return list(links)
def analyze(self, line):
if not line or line[0].startswith("#"):
return
date, _type, family, hostname, url, status, registrar, ips, asns, countries = tuple(
line)
tags = []
tags += TYPE_DICT[_type]
tags.append(family.lower())
context = {
"first_seen": date,
"status": status,
"registrar": registrar,
"countries": countries.split("|"),
"asns": asns.split("|"),
"source": self.name
}
try:
url = Url.get_or_create(value=url)
url.add_context(context)
url.tag(tags)
hostname = Observable.add_text(hostname)
hostname.tag(tags + ['blocklist'])
for ip in ips.split("|"):
if ip != hostname:
try:
i = Ip.get_or_create(value=ip)
i.active_link_to(hostname,
"First seen IP",
self.name,
clean_old=False)
except ObservableValidationError as e:
logging.error("Invalid Observable: {}".format(e))
except ObservableValidationError as e:
logging.error("Invalid line: {}\nLine: {}".format(e, line))
def _get_selected_observables(self, data):
if isinstance(data, MultiDict):
ids = data.getlist('ids')
query = data.get('query')
else:
ids = data.get('ids', None)
query = data.get('query', None)
if ids:
return Observable.objects(id__in=ids)
elif query:
query = json.loads(query)
fltr = query.get('filter', {})
params = query.get('params', {})
regex = params.pop('regex', False)
ignorecase = params.pop('ignorecase', False)
return get_queryset(Observable, fltr, regex, ignorecase)
else:
return []
def analyze(observable, results):
links = set()
params = {
'query': observable.value
}
data = PassiveTotalApi.get('/dns/passive', results.settings, params)
for record in data['results']:
first_seen = datetime.strptime(record['firstSeen'], "%Y-%m-%d %H:%M:%S")
last_seen = datetime.strptime(record['lastSeen'], "%Y-%m-%d %H:%M:%S")
new = Observable.add_text(record['resolve'])
if isinstance(observable, Hostname):
links.update(observable.link_to(new, "{} record".format(record['recordType']), 'PassiveTotal', first_seen, last_seen))
else:
links.update(new.link_to(observable, "{} record".format(record['recordType']), 'PassiveTotal', first_seen, last_seen))
return list(links)
def _get_selected_observables(self, data):
if isinstance(data, MultiDict):
ids = data.getlist("ids")
query = data.get("query")
else:
ids = data.get("ids", None)
query = data.get("query", None)
if ids:
return Observable.objects(id__in=ids)
elif query:
query = json.loads(query)
fltr = query.get("filter", {})
params = query.get("params", {})
regex = params.pop("regex", False)
ignorecase = params.pop("ignorecase", False)
return get_queryset(Observable, fltr, regex, ignorecase)
else:
return []
def _get_selected_observables(self, data):
if isinstance(data, MultiDict):
ids = data.getlist('ids')
query = data.get('query')
else:
ids = data.get('ids', None)
query = data.get('query', None)
if ids:
return Observable.objects(id__in=ids)
elif query:
query = json.loads(query)
fltr = query.get('filter', {})
params = query.get('params', {})
regex = params.pop('regex', False)
ignorecase = params.pop('ignorecase', False)
return get_queryset(Observable, fltr, regex, ignorecase)
else:
return []
def derive(observables):
"""Indicate that the module needs a specific attribute to work properly.
This function is only useful in abstract modules, in order to make sure
that modules that inherit from this class correctly defines needed class
attributes.
Args:
variables: a string or an array of strings containing the name of
needed class attributes.
Raises:
ModuleInitializationError: One of the needed attributes is not
correctly defined.
"""
new = []
observables = list(iterify(observables))
for i, observable in enumerate(observables):
try:
t = Observable.guess_type(observable)
temp = t(value=observable)
temp.clean()
observable = temp.value
observables[i] = observable
for a in analyzers.get(t, []):
new.extend([
n for n in a.analyze_string(observable)
if n and n not in observables
])
except ObservableValidationError:
pass
if len(new) == 0:
return observables
else:
return observables + derive(new)
def __add_attribute(self, instance, attribute, context, tags):
if attribute["category"] == "External analysis":
return
if attribute.get("type") in self.TYPES_TO_IMPORT:
context["id"] = attribute["event_id"]
context["link"] = urljoin(
self.instances[instance]["url"],
"/events/{}".format(attribute["event_id"]),
)
context["comment"] = attribute["comment"]
obs = Observable.add_text(attribute["value"])
if attribute["category"]:
obs.tag(attribute["category"].replace(" ", "_"))
if tags:
obs.tag(tags)
obs.add_context(context)
def info(self):
i = Observable.info(self)
i['mime_type'] = self.mime_type
i['hashes'] = self.hashes
i['filenames'] = self.filenames
return i
def info(self):
i = Observable.info(self)
i['mime_type'] = self.mime_type
i['hashes'] = self.hashes
return i
def match_observables(observables, save_matches=False, fetch_neighbors=True):
# Remove empty observables
observables = [refang(observable) for observable in observables if observable]
extended_query = set(observables) | set(derive(observables))
data = {
"matches": [],
"unknown": set(observables),
"entities": {},
"known": [],
"neighbors": [],
}
# add to "known"
for o in Observable.objects(value__in=list(extended_query)):
data['known'].append(o.info())
del_from_set(data['unknown'], o.value)
if fetch_neighbors:
for link, node in (o.incoming()):
if isinstance(node, Observable):
if (link.src.value not in extended_query or link.dst.value not in extended_query) and node.tags:
data['neighbors'].append((link.info(), node.info()))
for nodes in o.neighbors("Entity").values():
for l, node in nodes:
# add node name and link description to indicator
node_data = {"entity": node.type, "name": node.name, "link_description": l.description}
# uniquely add node information to related entitites
ent = data['entities'].get(node.name, node.info())
if 'matches' not in ent:
ent['matches'] = {"observables": []}
if 'observables' not in ent['matches']:
ent['matches']['observables'] = []
info = node.info()
o_info = o.info()
info['matched_observable'] = {
"value": o_info['value'],
"tags": [t['name'] for t in o_info['tags']],
"human_url": o_info['human_url'],
"url": o_info['url']
}
if info not in ent['matches']['observables']:
ent['matches']['observables'].append(info)
data['entities'][node.name] = ent
# add to "matches"
for o, i in Indicator.search(extended_query):
if save_matches:
o = Observable.add_text(o)
else:
o = Observable.guess_type(o)(value=o)
try:
o.validate()
except ObservableValidationError:
pass
try:
o = Observable.objects.get(value=o.value)
except Exception:
pass
match = i.info()
match.update({"observable": o.info(), "related": [], "suggested_tags": set()})
for nodes in i.neighbors("Entity").values():
for l, node in nodes:
# add node name and link description to indicator
node_data = {"entity": node.type, "name": node.name, "link_description": l.description}
match["related"].append(node_data)
# uniquely add node information to related entitites
ent = data['entities'].get(node.name, node.info())
if 'matches' not in ent:
ent['matches'] = {"indicators": []}
if 'indicators' not in ent['matches']:
ent['matches']['indicators'] = []
info = i.info()
info['matched_observable'] = o.value
if info not in ent['matches']['indicators']:
ent['matches']['indicators'].append(info)
data['entities'][node.name] = ent
o_tags = o.get_tags()
[match["suggested_tags"].add(tag) for tag in node.generate_tags() if tag not in o_tags]
data["matches"].append(match)
data['entities'] = data['entities'].values()
return data
et.save()
et = ExportTemplate(name="Bluecoat")
et.template = """define category cert_blocklist
{% for obs in elements %}{{ obs.value }}
{% endfor %}end
"""
et.save()
Export(name="TestExport",
acts_on="Url",
description="Test description",
frequency=timedelta(hours=1),
include_tags=[t1, t2],
template=et).save()
url = Observable.add_text("hxxp://zeuscpanel.com/gate.php")
url.tag(["zeus", "banker", "cc", "c2"])
print url.tags
# print url.find_tags()
# import pdb; pdb.set_trace()
## Create some instances of malware & co
bartalex = Malware.get_or_create(name="Bartalex")
bartalex.family = MalwareFamily.objects.get(name="dropper")
bartalex.killchain = "3"
bartalex.tags = ["bartalex"]
bartalex.save()
dridex = Malware.get_or_create(name="Dridex")
def each(module_name, observable_json):
o = Observable.from_json(observable_json)
mod = loaded_modules[module_name]
logging.debug("Launching {} on {}".format(mod.name, o))
mod.each(o)
o.analysis_done(mod.name)
MalwareFamily("rootkit").save()
MalwareFamily("trojan").save()
MalwareFamily("dropper").save()
t1 = Tag.get_or_create(name="zeus").add_produces(["crimeware", "banker", "malware"])
t2 = Tag.get_or_create(name="banker").add_produces(["crimeware", "malware"])
t3 = Tag.get_or_create(name="c2")
t3.add_replaces(["c&c", "cc"])
Tag.get_or_create(name="crimeware").add_produces("malware")
Export(name="TestExport", description="Test description", frequency=timedelta(hours=1), include_tags=[t1, t2]).save()
url = Observable.add_text("hxxp://zeuscpanel.com/gate.php")
url.tag(["zeus", "banker", "cc", "c2"])
print url.tags
# print url.find_tags()
# import pdb; pdb.set_trace()
## Create some instances of malware & co
bartalex = Malware.get_or_create(name="Bartalex")
bartalex.family = MalwareFamily.objects.get(name="dropper")
bartalex.killchain = "delivery"
bartalex.tags = ["bartalex"]
bartalex.save()
def match_observables(observables, save_matches=False, fetch_neighbors=True):
# Remove empty observables
observables, extended_query = derive(observables)
observables = list(observables)
data = {
"matches": [],
"unknown": set(observables),
"entities": {},
"known": [],
"neighbors": [],
}
# add to "known"
for o in Observable.objects(value__in=list(extended_query)):
data['known'].append(o.info())
del_from_set(data['unknown'], o.value)
if fetch_neighbors:
for link, node in (o.incoming()):
if isinstance(node, Observable):
if (link.src.value not in extended_query or link.dst.value
not in extended_query) and node.tags:
data['neighbors'].append((link.info(), node.info()))
for nodes in o.neighbors("Entity").values():
for l, node in nodes:
# add node name and link description to indicator
node_data = {
"entity": node.type,
"name": node.name,
"link_description": l.description
}
# uniquely add node information to related entitites
ent = data['entities'].get(node.name, node.info())
if 'matches' not in ent:
ent['matches'] = {"observables": []}
if 'observables' not in ent['matches']:
ent['matches']['observables'] = []
info = node.info()
o_info = o.info()
info['matched_observable'] = {
"value": o_info['value'],
"tags": [t['name'] for t in o_info['tags']],
"human_url": o_info['human_url'],
"url": o_info['url'],
"context": o_info['context']
}
if info not in ent['matches']['observables']:
ent['matches']['observables'].append(info)
data['entities'][node.name] = ent
# add to "matches"
for o, i in Indicator.search(extended_query):
if save_matches:
o = Observable.add_text(o)
else:
o = Observable.guess_type(o)(value=o)
try:
o.validate()
except ObservableValidationError:
pass
try:
o = Observable.objects.get(value=o.value)
except Exception:
pass
match = i.info()
match.update({
"observable": o.info() if o.id else o.value,
"related": [],
"suggested_tags": set()
})
for nodes in i.neighbors("Entity").values():
for l, node in nodes:
# add node name and link description to indicator
node_data = {
"entity": node.type,
"name": node.name,
"link_description": l.description
}
match["related"].append(node_data)
# uniquely add node information to related entitites
ent = data['entities'].get(node.name, node.info())
if 'matches' not in ent:
ent['matches'] = {"indicators": []}
if 'indicators' not in ent['matches']:
ent['matches']['indicators'] = []
info = i.info()
info['matched_observable'] = o.value
if info not in ent['matches']['indicators']:
ent['matches']['indicators'].append(info)
data['entities'][node.name] = ent
o_tags = o.get_tags()
for tag in node.generate_tags():
if tag not in o_tags:
match["suggested_tags"].add(tag)
data["matches"].append(match)
data['entities'] = data['entities'].values()
return data
def delete(self, id):
tag = self.objectmanager.objects.get(id=id)
tag.delete()
Observable.objects(tags__name=tag.name).update(pull__tags__name=tag.name)
return render({"status": "ok"})
def info(self):
i = Observable.info(self)
i["mime_type"] = self.mime_type
i["hashes"] = self.hashes
i["filenames"] = self.filenames
return i
def analyze(self, item):
first_seen = item["first_seen_utc"]
ioc_value = item["ioc_value"]
ioc_type = item["ioc_type"]
threat_type = item["threat_type"]
malware_alias = item["malware_alias"]
malware_printable = item["malware_printable"]
last_seen_utc = item["last_seen_utc"]
confidence_level = item["confidence_level"]
reference = item["reference"]
reporter = item["reporter"]
tags = []
context = {"source": self.name}
context["first_seen"] = first_seen
if reference:
context["reference"] = reference
else:
context["reference"] = "Unknown"
if reporter:
context["reporter"] = reporter
else:
context["reporter"] = "Unknown"
if threat_type:
context["threat_type"] = threat_type
if item["tags"]:
tags.extend(item["tags"].split(","))
if malware_printable:
tags.append(malware_printable)
if malware_alias:
context["malware_alias"] = malware_alias
if last_seen_utc:
context["last_seen_utc"] = last_seen_utc
if confidence_level:
context["confidence_level"] = confidence_level
value = None
obs = None
try:
if "ip" in ioc_type:
value, port = ioc_value.split(":")
context["port"] = port
obs = Ip.get_or_create(value=value)
else:
obs = Observable.add_text(ioc_value)
except ObservableValidationError as e:
logging.error(e)
return
if obs:
obs.add_context(context)
obs.add_source(self.name)
if tags:
obs.tag(tags)
if malware_printable:
obs.tags
def each(module_name, observable_json):
o = Observable.from_json(observable_json)
mod = ScheduledAnalytics.objects.get(name=module_name)
logging.debug("Launching {} on {}".format(mod.name, o))
mod.each(o)
o.analysis_done(mod.name)
import sys
import logging
from os import path
YETI_ROOT = path.normpath(path.dirname(path.dirname(path.abspath(__file__))))
sys.path.append(YETI_ROOT)
from core.analytics import OneShotAnalytics
from core.scheduling import Scheduler
from core.observables import Observable
if __name__ == '__main__':
logging.basicConfig(level=logging.DEBUG)
Scheduler()
if len(sys.argv) == 1:
print("Re-run using a analytic name as argument")
for f in OneShotAnalytics.objects():
print(" {}".format(f.name))
if len(sys.argv) > 1:
name = sys.argv[1]
f = OneShotAnalytics.objects.get(name=name)
print("Running {}...".format(f.name))
observable = Observable.guess_type(
sys.argv[2]).get_or_create(value=sys.argv[2])
f.analyze(observable, {})
def _make_threat_nodes(threat, context, tags):
# extract Url and Hash info
threats = dict()
if threat["threatStatus"] != "active":
# FIXME, clear out false positive ?
log.warning(
"threatStatus %s for threat %s",
threat["threatStatus"],
threat["threatID"],
)
log.debug(pprint.pformat(threat))
return None
log.debug("_make_threat_nodes for threat %s", threat["threatID"])
# threattype, classification
# url, phish: url leads to phishing page (threat is url)
# url, malware: url leads to malware download (threat is url, threatid is maybe sha256)
# attachment, malware: attachement is malware (threat is sha256)
# spam, url
if threat["threatType"] == "url":
if threat["classification"] == "phish":
pass # just keep the url
elif threat["classification"] == "malware":
# get url and hash
threats["attachment"] = threat
elif threat["classification"] == "spam":
log.info("URL threat - ignore classification %s",
threat["classification"])
else:
log.error("Type: url, Unsupported classification %s",
threat["classification"])
log.debug(pprint.pformat(threat))
return None
threats["url"] = threat
elif threat["threatType"] == "attachment":
if threat["classification"] == "malware":
threats["attachment"] = threat
else:
log.error(
"Type: attachment, Unsupported classification %s",
threat["classification"],
)
log.debug(pprint.pformat(threat))
return None
else:
log.error(
"Unsupported threatType %s classification %s",
threat["threatType"],
threat["classification"],
)
log.debug(pprint.pformat(threat))
return None
# FIXME check if they exist already.
# if they do, do not parse the threat a second time ?
threat_nodes = []
if "url" in threats:
# Proofpoint sometimes supplies a hostname marked as a Url.
# this relies on Yeti to determine the type/class and add act appropriately
threat_nodes.append(
Observable.guess_type(threats["url"]["threat"]).get_or_create(
value=threats["url"]["threat"], context=[context]))
if "attachment" in threats:
threat_nodes.append(
Hash.get_or_create(value=threats["attachment"]["threatID"],
context=[context]))
for o in threat_nodes:
o.tag([t["name"] for t in tags])
return threat_nodes
