feat(intel): persist per-provider taxonomy on AttackerIntel for TTP dispatch

The 2026-05-02 ship-time audit of the R0054-R0058 intel rule pack found
that AbuseIPDB / GreyNoise / ThreatFox stored only the aggregate verdict
(score / classification / listed-bool) plus the raw response blob. The
TTP IntelLifter expects per-provider taxonomy fields (categories, tags,
threat_types) that were never populated, so R0054 / R0055 / R0057
emitted zero tags in production despite passing unit tests.

Add typed columns: abuseipdb_categories, greynoise_tags, greynoise_name,
feodo_malware_family, threatfox_threat_types, threatfox_ioc_types,
threatfox_malware_families. Each provider now parses the relevant
taxonomy out of the upstream response and writes it through
column_updates. JSON-list columns ride as TEXT with default "[]" to
keep the SQLite/MySQL backend split honest, deserialised back to native
lists by the repo on read.
This commit is contained in:
2026-05-02 18:07:57 -04:00
parent d1c4a48963
commit 999d3494b4
10 changed files with 272 additions and 1 deletions

View File

@@ -95,6 +95,50 @@ async def test_low_score_maps_to_benign(monkeypatch):
assert result.column_updates["abuseipdb_score"] == 0
@pytest.mark.anyio
async def test_categories_flattened_from_reports(monkeypatch):
"""Post-2026-05-02 audit: provider must extract the union of
``data.reports[*].categories`` so the IntelLifter can dispatch
ATT&CK techniques. Sorted for deterministic test + bus diff."""
monkeypatch.setenv("DECNET_ABUSEIPDB_API_KEY", "k3y")
async def handler(request: httpx.Request) -> httpx.Response:
return httpx.Response(
200,
json={"data": {
"abuseConfidenceScore": 80,
"reports": [
{"categories": [18, 22]},
{"categories": [22, 14]},
{"categories": []},
{"not_a_dict": True},
{"categories": [21]},
],
}},
)
_install_transport(handler)
provider = AbuseIPDBProvider()
result = await provider.lookup("1.2.3.4")
cats = json.loads(result.column_updates["abuseipdb_categories"])
assert cats == [14, 18, 21, 22]
@pytest.mark.anyio
async def test_categories_empty_when_no_reports(monkeypatch):
monkeypatch.setenv("DECNET_ABUSEIPDB_API_KEY", "k3y")
async def handler(request: httpx.Request) -> httpx.Response:
return httpx.Response(
200, json={"data": {"abuseConfidenceScore": 5}},
)
_install_transport(handler)
provider = AbuseIPDBProvider()
result = await provider.lookup("8.8.8.8")
assert json.loads(result.column_updates["abuseipdb_categories"]) == []
@pytest.mark.anyio
async def test_429_returns_error(monkeypatch):
monkeypatch.setenv("DECNET_ABUSEIPDB_API_KEY", "k3y")

View File

@@ -87,6 +87,31 @@ async def test_unlisted_ip_returns_no_verdict():
assert result.column_updates["feodo_listed"] is False
@pytest.mark.anyio
async def test_listed_ip_persists_malware_family():
"""Post-2026-05-02 audit: IntelLifter reads
``feodo_malware_family`` for evidence; persist it as a typed
column rather than only inside ``feodo_raw``."""
async def handler(request: httpx.Request) -> httpx.Response:
return httpx.Response(200, json=_FEED)
_install_transport(handler)
provider = FeodoProvider(refresh_interval_s=999.0)
result = await provider.lookup("9.9.9.9")
assert result.column_updates["feodo_malware_family"] == "TrickBot"
@pytest.mark.anyio
async def test_unlisted_ip_clears_family():
async def handler(request: httpx.Request) -> httpx.Response:
return httpx.Response(200, json=_FEED)
_install_transport(handler)
provider = FeodoProvider(refresh_interval_s=999.0)
result = await provider.lookup("1.2.3.4")
assert result.column_updates["feodo_malware_family"] is None
@pytest.mark.anyio
async def test_feed_failure_reports_error():
async def handler(request: httpx.Request) -> httpx.Response:

View File

@@ -123,6 +123,45 @@ async def test_429_returns_error_no_writes():
assert result.column_updates == {}
@pytest.mark.anyio
async def test_actor_name_and_tags_persisted_when_present():
"""Post-2026-05-02 audit: ``name`` (actor label) and any ``tags``
list returned by the upstream survive into ``column_updates``.
The Community endpoint does not return ``tags`` in practice; the
test seeds the field anyway so non-Community provider plans that
do (paid / Enterprise) work without further code changes.
"""
async def handler(request: httpx.Request) -> httpx.Response:
return httpx.Response(
200,
json={
"classification": "malicious",
"name": "Tor",
"tags": ["tor_exit_node", "ssh_bruteforcer"],
},
)
provider = GreyNoiseProvider()
_install_transport(provider, handler)
result = await provider.lookup("1.2.3.4")
assert result.column_updates["greynoise_name"] == "Tor"
tags = json.loads(result.column_updates["greynoise_tags"])
assert tags == ["tor_exit_node", "ssh_bruteforcer"]
@pytest.mark.anyio
async def test_404_clears_actor_and_tags():
async def handler(request: httpx.Request) -> httpx.Response:
return httpx.Response(404, json={"message": "not seen"})
provider = GreyNoiseProvider()
_install_transport(provider, handler)
result = await provider.lookup("10.0.0.5")
assert result.column_updates["greynoise_name"] is None
assert result.column_updates["greynoise_tags"] == "[]"
@pytest.mark.anyio
async def test_network_failure_becomes_error():
async def handler(request: httpx.Request) -> httpx.Response:

View File

@@ -100,6 +100,61 @@ async def test_unexpected_status_is_error():
assert result.column_updates == {}
@pytest.mark.anyio
async def test_threat_types_and_ioc_types_flattened(monkeypatch):
"""Post-2026-05-02 audit: provider must extract the union of
``threat_type`` / ``ioc_type`` / ``malware`` across all matches.
The IntelLifter dispatches ATT&CK on threat_type."""
monkeypatch.delenv("DECNET_THREATFOX_API_KEY", raising=False)
async def handler(request: httpx.Request) -> httpx.Response:
return httpx.Response(
200,
json={"query_status": "ok", "data": [
{
"ioc_type": "ip:port",
"threat_type": "botnet_cc",
"malware": "Sliver",
},
{
"ioc_type": "url",
"threat_type": "payload_delivery",
"malware_printable": "Emotet",
},
{
"ioc_type": "ip:port", # duplicate, dedup'd
"threat_type": "botnet_cc", # duplicate
"malware": "Sliver", # duplicate
},
"not a dict — silently skipped",
]},
)
_install_transport(handler)
provider = ThreatFoxProvider()
result = await provider.lookup("1.2.3.4")
cu = result.column_updates
assert json.loads(cu["threatfox_threat_types"]) == [
"botnet_cc", "payload_delivery",
]
assert json.loads(cu["threatfox_ioc_types"]) == ["ip:port", "url"]
assert json.loads(cu["threatfox_malware_families"]) == ["Emotet", "Sliver"]
@pytest.mark.anyio
async def test_no_result_clears_taxonomy_columns():
async def handler(request: httpx.Request) -> httpx.Response:
return httpx.Response(200, json={"query_status": "no_result"})
_install_transport(handler)
provider = ThreatFoxProvider()
result = await provider.lookup("8.8.8.8")
cu = result.column_updates
assert cu["threatfox_threat_types"] == "[]"
assert cu["threatfox_ioc_types"] == "[]"
assert cu["threatfox_malware_families"] == "[]"
@pytest.mark.anyio
async def test_http_error_surfaces():
async def handler(request: httpx.Request) -> httpx.Response: