feat(creds): Phase 1 — Authorization header + SNMP community capture

Closes the cred-coverage gap for 7 services that already had the data
on the wire but never landed it in the Credential table:

- SNMP — community string lands as secret_kind="snmp_community",
  principal=None (v1/v2c has no per-user identity, the community IS
  the auth).
- SIP — Digest response hash, previously buried in the auth= header
  dump, now classify_authorization()-extracted.
- HTTP / HTTPS — Authorization header was in the headers JSON but
  never extracted. Now Basic decodes to plaintext, Bearer →
  http_bearer (principal=None), Digest → http_digest_md5.
- K8s — already extracted Authorization but didn't normalize. Service-
  account JWTs flow through as Bearer.
- Docker API — headers absent entirely. Adds the headers JSON dump
  and runs Authorization through the classifier.
- Elasticsearch — five distinct request handlers; each gains a
  per-handler _cred_fields() helper.

Adds canonical templates/syslog_bridge.py:classify_authorization().
Recognised: Basic / Bearer / Token / Digest. Unknown schemes (NTLM,
AWS4-HMAC, Negotiate) return None; the header still rides in the
ambient SD-block but isn't normalized as a credential. The SD shape
on the wire collapses sip_digest_md5 into http_digest_md5 — same
algorithm, so cross-protocol reuse correlates correctly when (rare)
nonce collisions allow.

Drive-by repair of tests/core/test_fingerprinting.py:

- The pre-existing `test_http_useragent_extracted` asserted both that
  add_bounty was called exactly once AND that the UA payload carried
  `path` and `method` fields. Both wrong since this session opened:
  the http_quirks fingerprint added later fires too, and the UA
  payload never actually included path/method despite the assertion.
- Adds `path`/`method` to the UA fingerprint payload (real operator
  value: "Nikto hit /admin" beats "Nikto seen on this decky").
- Replaces `assert_awaited_once` with a `_find_ua_bounty()` helper
  that filters add_bounty calls by `fingerprint_type`. New fingerprint
  families landing later won't retroactively break old tests.
- Updates the two credential-bearing tests to use the post-DEBT-039
  native shape (`secret_b64` / `principal`) and `upsert_credential`,
  not the deleted legacy `username+password` adapter.

Also rebuilds the per-service fake `syslog_bridge` modules in
tests/service_testing/{conftest,test_imap,test_pop3,test_snmp,test_mqtt,test_smtp}.py
to expose `encode_secret` + `classify_authorization`. Service templates
that import either now no longer fail at test collection.

173 tests pass in the touched scope. Phases 2-7 still pending.
This commit is contained in:
2026-04-25 07:04:10 -04:00
parent 6b16c844b6
commit 3404e3b3a6
44 changed files with 3081 additions and 66 deletions

View File

@@ -15,6 +15,23 @@ def _make_repo():
# HTTP User-Agent
# ---------------------------------------------------------------------------
def _find_ua_bounty(repo) -> dict:
"""Find the http_useragent fingerprint among all add_bounty calls.
A single HTTP request can produce multiple `bounty_type="fingerprint"`
bounties (UA, http_quirks, ip_leak, …). Tests for one specific kind
must filter rather than assert call count, so adding new fingerprint
families later doesn't retroactively break old tests."""
for c in repo.add_bounty.await_args_list:
payload = c[0][0].get("payload") or {}
if payload.get("fingerprint_type") == "http_useragent":
return c[0][0]
raise AssertionError(
"no http_useragent bounty found; calls=%r"
% [c[0][0].get("payload") for c in repo.add_bounty.await_args_list]
)
@pytest.mark.asyncio
async def test_http_useragent_extracted():
repo = _make_repo()
@@ -30,13 +47,12 @@ async def test_http_useragent_extracted():
},
}
await _extract_bounty(repo, log_data)
repo.add_bounty.assert_awaited_once()
call_kwargs = repo.add_bounty.call_args[0][0]
assert call_kwargs["bounty_type"] == "fingerprint"
assert call_kwargs["payload"]["fingerprint_type"] == "http_useragent"
assert call_kwargs["payload"]["value"] == "Nikto/2.1.6"
assert call_kwargs["payload"]["path"] == "/admin"
assert call_kwargs["payload"]["method"] == "GET"
bounty = _find_ua_bounty(repo)
assert bounty["bounty_type"] == "fingerprint"
assert bounty["payload"]["fingerprint_type"] == "http_useragent"
assert bounty["payload"]["value"] == "Nikto/2.1.6"
assert bounty["payload"]["path"] == "/admin"
assert bounty["payload"]["method"] == "GET"
@pytest.mark.asyncio
@@ -52,12 +68,14 @@ async def test_http_useragent_lowercase_key():
},
}
await _extract_bounty(repo, log_data)
call_kwargs = repo.add_bounty.call_args[0][0]
assert call_kwargs["payload"]["value"] == "sqlmap/1.7"
bounty = _find_ua_bounty(repo)
assert bounty["payload"]["value"] == "sqlmap/1.7"
@pytest.mark.asyncio
async def test_http_no_useragent_no_fingerprint_bounty():
"""No User-Agent header → no http_useragent bounty (other fingerprint
families like http_quirks may still fire on the same request)."""
repo = _make_repo()
log_data = {
"decky": "decky-01",
@@ -69,7 +87,11 @@ async def test_http_no_useragent_no_fingerprint_bounty():
},
}
await _extract_bounty(repo, log_data)
repo.add_bounty.assert_not_awaited()
ua_calls = [
c for c in repo.add_bounty.await_args_list
if (c[0][0].get("payload") or {}).get("fingerprint_type") == "http_useragent"
]
assert ua_calls == []
@pytest.mark.asyncio
@@ -142,39 +164,59 @@ async def test_vnc_version_event_no_client_version_field():
@pytest.mark.asyncio
async def test_credential_still_extracted_alongside_fingerprint():
"""Native-shape credential lands via upsert_credential, not add_bounty.
The legacy username+password adapter was deleted in DEBT-039; the
universal shape (secret_b64 + principal) goes straight to the
Credential table. Fingerprint bounties continue to ride add_bounty."""
import base64
repo = _make_repo()
repo.upsert_credential = AsyncMock()
log_data = {
"decky": "decky-03",
"service": "ftp",
"attacker_ip": "10.0.0.8",
"event_type": "auth_attempt",
"fields": {"username": "admin", "password": "1234"},
"fields": {
"username": "admin",
"principal": "admin",
"secret_kind": "plaintext",
"secret_printable": "1234",
"secret_b64": base64.b64encode(b"1234").decode(),
},
}
await _extract_bounty(repo, log_data)
repo.add_bounty.assert_awaited_once()
call_kwargs = repo.add_bounty.call_args[0][0]
assert call_kwargs["bounty_type"] == "credential"
repo.upsert_credential.assert_awaited_once()
cred = repo.upsert_credential.call_args[0][0]
assert cred["service"] == "ftp"
assert cred["principal"] == "admin"
@pytest.mark.asyncio
async def test_http_credential_and_fingerprint_both_extracted():
"""An HTTP login attempt can yield both a credential and a UA fingerprint."""
"""An HTTP login attempt yields both a Credential row and a UA
fingerprint bounty — distinct write paths."""
import base64
repo = _make_repo()
repo.upsert_credential = AsyncMock()
log_data = {
"decky": "decky-03",
"service": "http",
"attacker_ip": "10.0.0.9",
"event_type": "request",
"fields": {
"username": "root",
"password": "toor",
"principal": "root",
"secret_kind": "plaintext",
"secret_printable": "toor",
"secret_b64": base64.b64encode(b"toor").decode(),
"headers": {"User-Agent": "curl/7.88.1"},
},
}
await _extract_bounty(repo, log_data)
assert repo.add_bounty.await_count == 2
types = {c[0][0]["bounty_type"] for c in repo.add_bounty.call_args_list}
assert types == {"credential", "fingerprint"}
repo.upsert_credential.assert_awaited_once()
# add_bounty fired for the UA fingerprint; http_quirks may also fire.
bounty_types = {c[0][0]["bounty_type"] for c in repo.add_bounty.call_args_list}
assert "fingerprint" in bounty_types
# ---------------------------------------------------------------------------

View File

@@ -25,6 +25,14 @@ def make_fake_syslog_bridge() -> ModuleType:
mod.forward_syslog = MagicMock()
mod.SEVERITY_WARNING = 4
mod.SEVERITY_INFO = 6
# encode_secret returns the universal cred SD shape; tests don't
# care about the exact bytes, just that the key set is correct.
mod.encode_secret = MagicMock(
return_value={"secret_printable": "", "secret_b64": ""}
)
# classify_authorization returns None for unknown / absent auth so
# services that call **(cred or {}) get a no-op spread.
mod.classify_authorization = MagicMock(return_value=None)
return mod

View File

@@ -24,6 +24,8 @@ def _make_fake_syslog_bridge() -> ModuleType:
mod.forward_syslog = MagicMock()
mod.SEVERITY_WARNING = 4
mod.SEVERITY_INFO = 6
mod.encode_secret = MagicMock(return_value={"secret_printable": "", "secret_b64": ""})
mod.classify_authorization = MagicMock(return_value=None)
return mod

View File

@@ -23,6 +23,8 @@ def _make_fake_syslog_bridge() -> ModuleType:
mod.forward_syslog = MagicMock()
mod.SEVERITY_WARNING = 4
mod.SEVERITY_INFO = 6
mod.encode_secret = MagicMock(return_value={"secret_printable": "", "secret_b64": ""})
mod.classify_authorization = MagicMock(return_value=None)
return mod

View File

@@ -24,6 +24,8 @@ def _make_fake_syslog_bridge() -> ModuleType:
mod.forward_syslog = MagicMock()
mod.SEVERITY_WARNING = 4
mod.SEVERITY_INFO = 6
mod.encode_secret = MagicMock(return_value={"secret_printable": "", "secret_b64": ""})
mod.classify_authorization = MagicMock(return_value=None)
return mod

View File

@@ -27,6 +27,8 @@ def _make_fake_syslog_bridge() -> ModuleType:
mod.forward_syslog = MagicMock()
mod.SEVERITY_WARNING = 4
mod.SEVERITY_INFO = 6
mod.encode_secret = MagicMock(return_value={"secret_printable": "", "secret_b64": ""})
mod.classify_authorization = MagicMock(return_value=None)
return mod

View File

@@ -25,6 +25,8 @@ def _make_fake_syslog_bridge() -> ModuleType:
mod.forward_syslog = MagicMock()
mod.SEVERITY_WARNING = 4
mod.SEVERITY_INFO = 6
mod.encode_secret = MagicMock(return_value={"secret_printable": "", "secret_b64": ""})
mod.classify_authorization = MagicMock(return_value=None)
return mod

View File

@@ -222,6 +222,109 @@ async def test_vnc_hash_credential():
assert cred["secret_sha256"] == hashlib.sha256(raw).hexdigest()
@pytest.mark.asyncio
async def test_snmp_community_native_shape():
"""SNMP v1/v2c community string lands as secret_kind=snmp_community,
principal=None (no per-user identity in v1/v2c)."""
from decnet.web.ingester import _extract_bounty
repo = MagicMock(); repo.upsert_credential = AsyncMock()
raw = b"public"
log_data = {
"decky": "decky-01",
"service": "snmp",
"attacker_ip": "10.0.0.5",
"fields": {
"version": 1,
"community": "public",
"secret_kind": "snmp_community",
"secret_printable": "public",
"secret_b64": base64.b64encode(raw).decode("ascii"),
},
}
await _extract_bounty(repo, log_data)
cred = repo.upsert_credential.call_args[0][0]
assert cred["service"] == "snmp"
assert cred["secret_kind"] == "snmp_community"
assert cred["principal"] is None
assert cred["secret_sha256"] == hashlib.sha256(raw).hexdigest()
@pytest.mark.asyncio
async def test_http_basic_native_shape():
"""HTTP Basic via classify_authorization → principal+plaintext."""
from decnet.web.ingester import _extract_bounty
repo = MagicMock(); repo.upsert_credential = AsyncMock()
log_data = {
"decky": "decky-01",
"service": "http",
"attacker_ip": "10.0.0.5",
"fields": {
"method": "GET",
"path": "/admin",
"principal": "admin",
"secret_kind": "plaintext",
"secret_printable": "hunter2",
"secret_b64": base64.b64encode(b"hunter2").decode("ascii"),
},
}
await _extract_bounty(repo, log_data)
cred = repo.upsert_credential.call_args[0][0]
assert cred["service"] == "http"
assert cred["principal"] == "admin"
assert cred["secret_kind"] == "plaintext"
@pytest.mark.asyncio
async def test_http_bearer_native_shape():
"""HTTP Bearer — principal=None, secret_kind=http_bearer, opaque."""
from decnet.web.ingester import _extract_bounty
repo = MagicMock(); repo.upsert_credential = AsyncMock()
token = b"eyJhbGciOiJIUzI1NiJ9.foo.bar"
log_data = {
"decky": "decky-01",
"service": "k8s",
"attacker_ip": "10.0.0.5",
"fields": {
"method": "GET",
"path": "/api/v1/secrets",
"principal": None,
"secret_kind": "http_bearer",
"secret_printable": token.decode(),
"secret_b64": base64.b64encode(token).decode("ascii"),
},
}
await _extract_bounty(repo, log_data)
cred = repo.upsert_credential.call_args[0][0]
assert cred["secret_kind"] == "http_bearer"
assert cred["principal"] is None
assert cred["secret_sha256"] == hashlib.sha256(token).hexdigest()
@pytest.mark.asyncio
async def test_sip_digest_native_shape():
"""SIP Digest via classify_authorization → response hash captured."""
from decnet.web.ingester import _extract_bounty
repo = MagicMock(); repo.upsert_credential = AsyncMock()
response_hash = "d41d8cd98f00b204e9800998ecf8427e"
log_data = {
"decky": "decky-01",
"service": "sip",
"attacker_ip": "10.0.0.5",
"fields": {
"method": "REGISTER",
"principal": "alice",
"secret_kind": "http_digest_md5",
"secret_printable": response_hash,
"secret_b64": base64.b64encode(response_hash.encode()).decode("ascii"),
},
}
await _extract_bounty(repo, log_data)
cred = repo.upsert_credential.call_args[0][0]
assert cred["service"] == "sip"
assert cred["secret_kind"] == "http_digest_md5"
assert cred["principal"] == "alice"
@pytest.mark.asyncio
async def test_lossless_b64_survives_nonprintable_password():
"""Even when secret_printable is sanitized, secret_b64 still decodes

View File

@@ -65,6 +65,53 @@ def test_encode_secret_preserves_rfc5424_specials(syslog_bridge):
assert base64.b64decode(out["secret_b64"]) == secret.encode("utf-8")
def test_classify_authorization_basic(syslog_bridge):
"""HTTP Basic — base64(user:pw) decodes to plaintext credential."""
cred = syslog_bridge.classify_authorization("Basic YWRtaW46aHVudGVyMg==")
assert cred is not None
assert cred["principal"] == "admin"
assert cred["secret_kind"] == "plaintext"
assert base64.b64decode(cred["secret_b64"]) == b"hunter2"
assert cred["secret_printable"] == "hunter2"
def test_classify_authorization_bearer(syslog_bridge):
cred = syslog_bridge.classify_authorization("Bearer eyJhbGciOiJIUzI1NiJ9.foo.bar")
assert cred["principal"] is None
assert cred["secret_kind"] == "http_bearer"
assert base64.b64decode(cred["secret_b64"]) == b"eyJhbGciOiJIUzI1NiJ9.foo.bar"
def test_classify_authorization_token_alias(syslog_bridge):
"""`Token <opaque>` = same shape as Bearer (Kubernetes service accounts)."""
cred = syslog_bridge.classify_authorization("Token sa-jwt-token-abc")
assert cred["secret_kind"] == "http_bearer"
def test_classify_authorization_digest(syslog_bridge):
"""RFC 7616 Digest — extract username + response hash."""
header = ('Digest username="alice", realm="example.com", '
'nonce="abc123", uri="/", response="d41d8cd98f00b204e9800998ecf8427e"')
cred = syslog_bridge.classify_authorization(header)
assert cred["principal"] == "alice"
assert cred["secret_kind"] == "http_digest_md5"
assert cred["secret_printable"] == "d41d8cd98f00b204e9800998ecf8427e"
def test_classify_authorization_unknown_scheme(syslog_bridge):
"""NTLM, AWS4-HMAC-…, Negotiate — all return None for now."""
assert syslog_bridge.classify_authorization("NTLM TlRMTVNTUAA=") is None
assert syslog_bridge.classify_authorization("AWS4-HMAC-SHA256 Credential=…") is None
def test_classify_authorization_malformed(syslog_bridge):
assert syslog_bridge.classify_authorization(None) is None
assert syslog_bridge.classify_authorization("") is None
assert syslog_bridge.classify_authorization("Basic !!not-base64!!") is None
assert syslog_bridge.classify_authorization("Basic dXNlcg==") is None # no colon
assert syslog_bridge.classify_authorization("Digest no-response-here") is None
def test_encode_secret_unicode_replaced(syslog_bridge):
"""Non-ASCII unicode encodes via utf-8, then printable strips the
multi-byte sequence to '?' chars (one per raw byte)."""