feat(prober-cert): roll up fingerprints onto AttackerIdentity

Brings the federation-gossip columns on AttackerIdentity to life —
ja3_hashes, hassh_hashes, and the new tls_cert_sha256 — by projecting
the union of every member observation's fingerprints JSON onto the
identity at clusterer create / link / merge time.

- decnet/profiler/identity_rollup.py: pure extract_fp_summaries()
  reads the production bounty shape (payload.fingerprint_type +
  payload.{ja3,hash,cert_sha256}) and returns deduped+sorted JSON
  list[str] per family, or None when a family has no signal so the
  column stays NULL instead of '[]'.
- BaseRepository.update_identity_fingerprints + SQLModel impl: one
  idempotent write that overwrites the three summary columns and
  bumps updated_at.
- ConnectedComponentsClusterer: after every per-component
  reconciliation (fresh-create OR existing-merge+link), recomputes
  and writes the rollup for the target identity. Wrapped in a
  best-effort helper so a write failure logs but never breaks the
  tick.
- Tests: extract_fp_summaries unit (dedup, sort determinism,
  unknown types ignored, malformed JSON, nested-stringified
  payloads, non-string values); end-to-end clusterer ticks
  populate the columns on create + on later observation links;
  no-fingerprint clusters keep the columns NULL.

tests/clustering/test_connected_components.py

@@ -140,14 +140,41 @@ async def repo(tmp_path):
 
 async def _seed_attacker(
     repo, ip: str, *,
-    ja3: str | None = None, hassh: str | None = None, asn: int | None = None,
+    ja3: str | None = None,
+    hassh: str | None = None,
+    asn: int | None = None,
+    cert_sha256: str | None = None,
 ) -> str:
     now = datetime.now(timezone.utc)
-    fingerprints = []
+    # Two-shape fingerprint payload:
+    #   - the "kind" entries feed the clusterer's from_attacker_row
+    #     (test-fixture shape, line ~115 of connected_components.py)
+    #   - the "bounty_type/payload" entries feed identity_rollup's
+    #     extract_fp_summaries (production shape, written by the
+    #     profiler from real bounty rows). Both shapes coexist in
+    #     the same JSON list so the same seed exercises clustering
+    #     AND the identity-column rollup.
+    fingerprints: list[dict] = []
     if ja3:
         fingerprints.append({"kind": "ja3", "hash": ja3})
+        fingerprints.append({
+            "bounty_type": "fingerprint",
+            "payload": {"fingerprint_type": "ja3", "ja3": ja3},
+        })
     if hassh:
         fingerprints.append({"kind": "hassh", "hash": hassh})
+        fingerprints.append({
+            "bounty_type": "fingerprint",
+            "payload": {"fingerprint_type": "hassh_server", "hash": hassh},
+        })
+    if cert_sha256:
+        fingerprints.append({
+            "bounty_type": "fingerprint",
+            "payload": {
+                "fingerprint_type": "tls_certificate",
+                "cert_sha256": cert_sha256,
+            },
+        })
     return await repo.upsert_attacker({
         "ip": ip,
         "first_seen": now,
@@ -377,6 +404,70 @@ async def test_tick_links_new_observation_to_existing_identity(repo):
     assert d in linked_uuids
 
 
+# ─── identity fingerprint rollup ───────────────────────────────────────────
+
+
+@pytest.mark.anyio
+async def test_tick_rolls_up_fingerprint_columns_on_create(repo):
+    """A fresh-component tick must populate ja3_hashes / hassh_hashes /
+    tls_cert_sha256 on the newly-minted identity row, deduplicated and
+    sorted across all member observations."""
+    await _seed_attacker(
+        repo, "1.1.1.1", ja3="ja3-x", hassh="hassh-y", cert_sha256="ab" * 32,
+    )
+    await _seed_attacker(
+        repo, "2.2.2.2", ja3="ja3-x", hassh="hassh-y", cert_sha256="cd" * 32,
+    )
+    c = ConnectedComponentsClusterer()
+    result = await c.tick(repo)
+    assert len(result.identities_formed) == 1
+    identity_uuid = result.identities_formed[0]["identity_uuid"]
+
+    rows = {i["uuid"]: i for i in await repo.list_all_identities()}
+    identity = rows[identity_uuid]
+    assert json.loads(identity["ja3_hashes"]) == ["ja3-x"]
+    assert json.loads(identity["hassh_hashes"]) == ["hassh-y"]
+    assert json.loads(identity["tls_cert_sha256"]) == sorted(["ab" * 32, "cd" * 32])
+
+
+@pytest.mark.anyio
+async def test_tick_rolls_up_fingerprints_on_link(repo):
+    """When a new observation links into an existing identity, the
+    rollup must reflect any new cert SHA-256 it brings."""
+    await _seed_attacker(
+        repo, "1.1.1.1", ja3="ja3-x", cert_sha256="ab" * 32,
+    )
+    c = ConnectedComponentsClusterer()
+    first = await c.tick(repo)
+    identity_uuid = first.identities_formed[0]["identity_uuid"]
+
+    # New observation, same JA3, fresh cert.
+    await _seed_attacker(
+        repo, "2.2.2.2", ja3="ja3-x", cert_sha256="cd" * 32,
+    )
+    await c.tick(repo)
+
+    rows = {i["uuid"]: i for i in await repo.list_all_identities()}
+    identity = rows[identity_uuid]
+    assert json.loads(identity["tls_cert_sha256"]) == sorted(["ab" * 32, "cd" * 32])
+
+
+@pytest.mark.anyio
+async def test_tick_leaves_columns_null_when_no_fingerprints(repo):
+    """Two attackers with NO fingerprint signal cluster as separate
+    singletons; their identity rows must keep all rollup columns NULL
+    (not "[]" — NULL distinguishes 'no signal yet' from 'known empty')."""
+    await _seed_attacker(repo, "1.1.1.1")
+    await _seed_attacker(repo, "2.2.2.2")
+    c = ConnectedComponentsClusterer()
+    await c.tick(repo)
+
+    for identity in await repo.list_all_identities():
+        assert identity["ja3_hashes"] is None
+        assert identity["hassh_hashes"] is None
+        assert identity["tls_cert_sha256"] is None
+
+
 # ─── fixture-bound assertions (in-memory) ──────────────────────────────────
 
 

tests/profiler/test_identity_rollup.py

@@ -0,0 +1,141 @@
+"""Tests for ``decnet.profiler.identity_rollup.extract_fp_summaries``.
+
+Pure unit tests against the production bounty shape that
+``decnet.profiler.worker._build_record`` writes into
+``Attacker.fingerprints`` — a list of ``{bounty_type, payload, ...}``
+dicts where the meaningful data lives under ``payload.fingerprint_type``.
+"""
+
+from __future__ import annotations
+
+import json
+
+from decnet.profiler.identity_rollup import extract_fp_summaries
+
+
+def _bounty(fp_type: str, **payload_extras) -> dict:
+    """Build a bounty dict shaped the way the profiler writes it."""
+    return {
+        "bounty_type": "fingerprint",
+        "payload": {"fingerprint_type": fp_type, **payload_extras},
+    }
+
+
+def _row_with(*entries) -> dict:
+    return {"fingerprints": json.dumps(list(entries))}
+
+
+class TestExtractFpSummaries:
+
+    def test_empty_input_returns_all_none(self):
+        result = extract_fp_summaries([])
+        assert result == {
+            "ja3_hashes": None,
+            "hassh_hashes": None,
+            "tls_cert_sha256": None,
+        }
+
+    def test_single_row_single_cert(self):
+        row = _row_with(_bounty("tls_certificate", cert_sha256="ab" * 32))
+        result = extract_fp_summaries([row])
+        assert result["ja3_hashes"] is None
+        assert result["hassh_hashes"] is None
+        assert json.loads(result["tls_cert_sha256"]) == ["ab" * 32]
+
+    def test_dedupe_across_rows(self):
+        sha = "ab" * 32
+        a = _row_with(_bounty("tls_certificate", cert_sha256=sha))
+        b = _row_with(_bounty("tls_certificate", cert_sha256=sha))
+        result = extract_fp_summaries([a, b])
+        assert json.loads(result["tls_cert_sha256"]) == [sha]
+
+    def test_sorted_output_is_deterministic(self):
+        a = _row_with(
+            _bounty("tls_certificate", cert_sha256="ff" * 32),
+            _bounty("tls_certificate", cert_sha256="11" * 32),
+            _bounty("tls_certificate", cert_sha256="aa" * 32),
+        )
+        result = extract_fp_summaries([a])
+        # Same input twice must produce byte-identical output.
+        assert result == extract_fp_summaries([a])
+        assert json.loads(result["tls_cert_sha256"]) == sorted(
+            ["ff" * 32, "11" * 32, "aa" * 32]
+        )
+
+    def test_all_three_families_at_once(self):
+        row = _row_with(
+            _bounty("ja3", ja3="ja3-abc"),
+            _bounty("hassh_server", hash="hassh-def"),
+            _bounty("tls_certificate", cert_sha256="ab" * 32),
+        )
+        result = extract_fp_summaries([row])
+        assert json.loads(result["ja3_hashes"]) == ["ja3-abc"]
+        assert json.loads(result["hassh_hashes"]) == ["hassh-def"]
+        assert json.loads(result["tls_cert_sha256"]) == ["ab" * 32]
+
+    def test_unknown_fingerprint_type_ignored(self):
+        # tcpfp / ja4l / http_quirks have no rollup column yet; they
+        # must not pollute the three families that do.
+        row = _row_with(
+            _bounty("tcpfp", hash="tcpfp-x"),
+            _bounty("ja4l", ja4l="ja4l-y"),
+            _bounty("http_quirks", quirks="..."),
+        )
+        result = extract_fp_summaries([row])
+        assert result["ja3_hashes"] is None
+        assert result["hassh_hashes"] is None
+        assert result["tls_cert_sha256"] is None
+
+    def test_missing_payload_key_skipped(self):
+        # tls_certificate bounty shaped like a sniffer-only payload
+        # (no cert_sha256). Must not crash, must not record an entry.
+        row = _row_with({
+            "bounty_type": "fingerprint",
+            "payload": {"fingerprint_type": "tls_certificate", "subject_cn": "x"},
+        })
+        result = extract_fp_summaries([row])
+        assert result["tls_cert_sha256"] is None
+
+    def test_malformed_fingerprints_json_returns_all_none(self):
+        result = extract_fp_summaries([{"fingerprints": "not json"}])
+        assert all(v is None for v in result.values())
+
+    def test_missing_fingerprints_field_returns_all_none(self):
+        result = extract_fp_summaries([{"some_other_field": True}])
+        assert all(v is None for v in result.values())
+
+    def test_payload_as_string_is_json_decoded(self):
+        # Defensive: some legacy storage may have nested-stringified payloads.
+        row = {
+            "fingerprints": json.dumps([{
+                "bounty_type": "fingerprint",
+                "payload": json.dumps({
+                    "fingerprint_type": "tls_certificate",
+                    "cert_sha256": "cd" * 32,
+                }),
+            }]),
+        }
+        result = extract_fp_summaries([row])
+        assert json.loads(result["tls_cert_sha256"]) == ["cd" * 32]
+
+    def test_non_string_hash_values_skipped(self):
+        row = _row_with({
+            "bounty_type": "fingerprint",
+            "payload": {"fingerprint_type": "tls_certificate", "cert_sha256": 12345},
+        })
+        result = extract_fp_summaries([row])
+        assert result["tls_cert_sha256"] is None
+
+    def test_dedup_across_many_rows_with_overlap(self):
+        rows = [
+            _row_with(_bounty("ja3", ja3="ja3-shared")),
+            _row_with(
+                _bounty("ja3", ja3="ja3-shared"),
+                _bounty("ja3", ja3="ja3-second"),
+            ),
+            _row_with(_bounty("ja3", ja3="ja3-third")),
+        ]
+        result = extract_fp_summaries(rows)
+        assert json.loads(result["ja3_hashes"]) == sorted(
+            ["ja3-shared", "ja3-second", "ja3-third"]
+        )