Commit Graph

7 Commits

Author SHA1 Message Date
24cdef9246 feat(smtp_relay): ingest probe_forwarded as probe_relay bounty
Adds probe_forwarded to meaningful event kinds and stores it in the
bounty table as bounty_type=probe_relay with forwarded=true/false, so
the dashboard shows whether the upstream actually accepted the test email.
2026-04-30 11:32:14 -04:00
4b7cb42ab1 fix(profiler): extract commands when MSGID=command, not just MSGID=NIL
The Dockerfile PROMPT_COMMAND logger uses --msgid command, so the MSGID
field arrives as 'command' not '-'. The CMD rewrite block was guarded by
event_type == '-' so it never fired, leaving fields['command'] unpopulated
and cmd_text=None for every SSH session command.

Broaden the guard to also match event_type == 'command' with no existing
'command' field, which covers both the intended (MSGID=NIL) and actual
(MSGID=command) wire formats.
2026-04-30 10:57:29 -04:00
d4591b38dc fix(profiler): aggregate bash PROMPT_COMMAND lines into attacker profile
SSH/telnet decky containers emit shell commands via `logger -t bash "CMD …"`
which produces RFC 5424 lines with MSGID=NIL. Both parsers were leaving
event_type="-", so the behavioral profiler's `_COMMAND_EVENT_TYPES` filter
silently dropped them — the IP profile existed but no command transcripts
or artifacts. Confirmed in the wild: 44/48 events from one attacker were
event_type="-".

Rewrite event_type to "command" in both parsers when MSGID=NIL and the
msg starts with "CMD ". Correlation parser also extracts the cmd= payload
into fields["command"] so the profiler can build the transcript; collector
parser leaves fields={} to avoid duplicate pills in the dashboard.
2026-04-28 19:09:41 -04:00
862e4dbb31 merge: testing → main (reconcile 2-week divergence) 2026-04-28 18:36:00 -04:00
b2e4706a14 Refactor: implemented Repository Factory and Async Mutator Engine. Decoupled storage logic and enforced Dependency Injection across CLI and Web API. Updated documentation.
Some checks failed
CI / Lint (ruff) (push) Successful in 12s
CI / SAST (bandit) (push) Successful in 13s
CI / Dependency audit (pip-audit) (push) Successful in 22s
CI / Test (Standard) (3.11) (push) Failing after 54s
CI / Test (Standard) (3.12) (push) Successful in 1m35s
CI / Test (Live) (3.11) (push) Has been skipped
CI / Test (Fuzz) (3.11) (push) Has been skipped
CI / Merge dev → testing (push) Has been skipped
CI / Prepare Merge to Main (push) Has been skipped
CI / Finalize Merge to Main (push) Has been skipped
2026-04-12 07:48:17 -04:00
988732f4f9 Fix all ruff lint errors across decnet/, templates/, and tests/
Some checks failed
CI / Test (pytest) (3.11) (push) Has been cancelled
CI / Test (pytest) (3.12) (push) Has been cancelled
Security / SAST (bandit) (push) Has been cancelled
Security / Dependency audit (pip-audit) (push) Has been cancelled
CI / Lint (ruff) (push) Has been cancelled
2026-04-04 17:36:16 -03:00
bff03d1198 Add cross-decky correlation engine and decnet correlate command
When the same attacker IP touches multiple deckies, the engine builds a
chronological traversal graph and reports the lateral movement path.

decnet/correlation/
  parser.py   — RFC 5424 line → LogEvent; handles src_ip + src field variants
  graph.py    — AttackerTraversal / TraversalHop data types with path/duration
  engine.py   — CorrelationEngine: ingest(), traversals(), report_table/json,
                traversal_syslog_lines() (emits WARNING-severity RFC 5424)
  __init__.py — public API re-exports

decnet/cli.py — `decnet correlate` command (--log-file, --min-deckies,
                --output table|json|syslog, --emit-syslog)

tests/test_correlation.py — 49 tests: parser, graph, engine, reporting

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-04 13:53:30 -03:00