classifyre-cli 0.4.15__tar.gz → 0.4.16__tar.gz

{classifyre_cli-0.4.15 → classifyre_cli-0.4.16}/.turbo/turbo-build.log

@@ -1,3 +1,3 @@
 $ uv sync
-Resolved 254 packages in 157ms
+Resolved 255 packages in 216ms
 Checked 50 packages in 1ms

{classifyre_cli-0.4.15 → classifyre_cli-0.4.16}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: classifyre-cli
-Version: 0.4.15
+Version: 0.4.16
 Summary: Classifyre CLI — scan and classify unstructured data sources
 License: MIT
 Keywords: data,ingestion,metadata,pii,secrets,unstructured

{classifyre_cli-0.4.15 → classifyre_cli-0.4.16}/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@classifyre/cli",
-  "version": "0.4.15",
+  "version": "0.4.16",
   "private": true,
   "scripts": {
     "build": "uv sync",

{classifyre_cli-0.4.15 → classifyre_cli-0.4.16}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "classifyre-cli"
-version = "0.4.15"
+version = "0.4.16"
 description = "Classifyre CLI — scan and classify unstructured data sources"
 readme = "README.md"
 requires-python = ">=3.12"
@@ -94,6 +94,7 @@ custom = [
     "transformers>=4.50.0,<6.0.0",
     "torch>=2.12.0",
     "pillow>=12.2.0",
+    "timm>=1.0.27",
     "setfit>=1.1.3",
     "datasets>=4.8.5",
     "scikit-learn>=1.7.2",

{classifyre_cli-0.4.15 → classifyre_cli-0.4.16}/src/detectors/custom/runners/_object_detection.py

@@ -48,6 +48,7 @@ class ObjectDetectionRunner(BaseRunner):
             raise MissingDependencyError(
                 "object_detection",
                 ["custom", "detectors"],
+                ["custom", "detectors"],
                 f"ObjectDetectionRunner requires additional dependencies: {exc}",
             ) from exc
 

{classifyre_cli-0.4.15 → classifyre_cli-0.4.16}/src/main.py

@@ -372,6 +372,16 @@ async def run_command_async(args: argparse.Namespace, recipe: dict[str, Any]) ->
                         total_assets,
                         output_batch_count,
                     )
+
+                    # Phase 1: emit source-derived relationship edges (best-effort).
+                    if hasattr(source, "collect_relationships") and hasattr(sink, "emit_edges"):
+                        try:
+                            edges = await source.collect_relationships()
+                            if edges:
+                                await sink.emit_edges(edges)
+                                logger.info("Emitted %d source-derived relationship edges", len(edges))
+                        except Exception as rel_error:
+                            logger.warning("Relationship emission failed (non-fatal): %s", rel_error)
                 except Exception as extraction_error:
                     if _is_timeout_error(extraction_error):
                         logger.warning(

{classifyre_cli-0.4.15 → classifyre_cli-0.4.16}/src/outputs/rest.py

@@ -97,6 +97,31 @@ class BulkIngestAssetsRequest(BaseModel):
     skip_findings: bool = Field(False, serialization_alias="skipFindings")
 
 
+class IngestEdge(BaseModel):
+    """A source-derived relationship edge for the investigation graph.
+
+    Identify endpoints by UUID (from_id / to_id) or by asset hash
+    (from_hash / to_hash — the API resolves hashes to UUIDs).
+    """
+
+    model_config = ConfigDict(populate_by_name=True)
+
+    from_type: str = Field(serialization_alias="fromType")
+    from_id: str | None = Field(None, serialization_alias="fromId")
+    from_hash: str | None = Field(None, serialization_alias="fromHash")
+    to_type: str = Field(serialization_alias="toType")
+    to_id: str | None = Field(None, serialization_alias="toId")
+    to_hash: str | None = Field(None, serialization_alias="toHash")
+    relation_type: str = Field(serialization_alias="relationType")
+    confidence: float = 1.0
+
+
+class BulkIngestEdgesRequest(BaseModel):
+    model_config = ConfigDict(populate_by_name=True)
+
+    edges: list[IngestEdge]
+
+
 class FinalizeIngestRunRequest(BaseModel):
     model_config = ConfigDict(populate_by_name=True)
 
@@ -252,6 +277,30 @@ class RestOutputSink:
                 update_error,
             )
 
+    async def emit_edges(self, edges: list[IngestEdge]) -> None:
+        """Bulk-upsert source-derived relationship edges to the investigation graph.
+
+        Idempotent — safe to call multiple times with overlapping data.
+        Silently skips if the list is empty.
+        """
+        if not edges:
+            return
+
+        _edge_batch = 500
+        for i in range(0, len(edges), _edge_batch):
+            chunk = edges[i : i + _edge_batch]
+            payload = BulkIngestEdgesRequest(edges=chunk)
+            try:
+                self._request_json(
+                    "POST",
+                    "/graph/edges",
+                    payload.model_dump(mode="json", by_alias=True),
+                )
+                logger.debug("Emitted %d source-derived edges to graph", len(chunk))
+            except Exception as exc:
+                # Edge emission is best-effort: log and continue.
+                logger.warning("Failed to emit edges to graph: %s", exc)
+
     async def register_discovered_assets(self, hashes: list[str]) -> None:
         runner_id = self._require_runner_id()
         for i in range(0, len(hashes), 500):

{classifyre_cli-0.4.15 → classifyre_cli-0.4.16}/src/sources/base.py

@@ -4,6 +4,7 @@ from collections.abc import AsyncGenerator, Generator
 from typing import TYPE_CHECKING, Any
 
 from ..models.generated_single_asset_scan_results import DetectionResult, SingleAssetScanResults
+from ..outputs.rest import IngestEdge
 
 if TYPE_CHECKING:
     from ..utils.file_parser import ParsedBytes
@@ -315,3 +316,15 @@ class BaseSource(ABC):
         this and map those identifiers back to their original URLs.
         """
         return normalize_http_url(link)
+
+
+    async def collect_relationships(self) -> list[IngestEdge]:
+        """Return source-derived relationship edges for the investigation graph.
+
+        Connectors override this to emit typed edges (READS, ATTACHED_TO,
+        SENT_TO, OWNS, ACCESSED, etc.) discovered during extraction. The caller
+        (main.py) will forward these to ``RestOutputSink.emit_edges()``.
+
+        Default: no relationships (empty list).
+        """
+        return []

{classifyre_cli-0.4.15 → classifyre_cli-0.4.16}/src/sources/email/source.py

@@ -102,6 +102,8 @@ class EmailSource(BaseSource):
         self._email_locator_by_hash: dict[str, tuple[str, str]] = {}
         self._hash_by_url: dict[str, str] = {}
         self._attachment_name_by_hash = {}
+        # Phase 1: tracks (email_hash → [attachment_hash, ...]) for relationship emission.
+        self._email_attachment_links: dict[str, list[str]] = {}
 
         self._mailbox: Any = None
 
@@ -303,6 +305,9 @@ class EmailSource(BaseSource):
             runner_id=self.runner_id,
             **self.metadata_fields("email", metadata),
         )
+        # Track for Phase 1 relationship emission.
+        if attachment_hashes:
+            self._email_attachment_links[email_hash] = attachment_hashes
         return [email_asset, *attachment_assets]
 
     def _attachment_to_asset(
@@ -515,6 +520,24 @@ class EmailSource(BaseSource):
     # ------------------------------------------------------------------
     # Lifecycle
     # ------------------------------------------------------------------
+    async def collect_relationships(self) -> list[Any]:
+        """Emit ATTACHED_TO edges: email → attachment for each extracted message."""
+        from ...outputs.rest import IngestEdge
+
+        edges: list[IngestEdge] = []
+        for email_hash, attachment_hashes in self._email_attachment_links.items():
+            for att_hash in attachment_hashes:
+                edges.append(
+                    IngestEdge(
+                        from_type="asset",
+                        from_hash=email_hash,
+                        to_type="asset",
+                        to_hash=att_hash,
+                        relation_type="ATTACHED_TO",
+                    )
+                )
+        return edges
+
     def abort(self) -> None:
         logger.info("Aborting email extraction...")
         super().abort()

{classifyre_cli-0.4.15 → classifyre_cli-0.4.16}/uv.lock

@@ -606,7 +606,7 @@ wheels = [
 
 [[package]]
 name = "classifyre-cli"
-version = "0.4.15"
+version = "0.4.16"
 source = { editable = "." }
 dependencies = [
     { name = "beautifulsoup4" },
@@ -643,6 +643,7 @@ custom = [
     { name = "scikit-learn" },
     { name = "sentence-transformers" },
     { name = "setfit" },
+    { name = "timm" },
     { name = "torch", version = "2.12.0", source = { registry = "https://pypi.org/simple" }, marker = "sys_platform != 'linux'" },
     { name = "torch", version = "2.12.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "sys_platform == 'linux'" },
     { name = "transformers" },
@@ -678,6 +679,7 @@ detectors = [
     { name = "setfit" },
     { name = "spacy" },
     { name = "textstat" },
+    { name = "timm" },
     { name = "torch", version = "2.12.0", source = { registry = "https://pypi.org/simple" }, marker = "sys_platform != 'linux'" },
     { name = "torch", version = "2.12.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "sys_platform == 'linux'" },
     { name = "transformers" },
@@ -821,6 +823,7 @@ custom = [
     { name = "scikit-learn", specifier = ">=1.7.2" },
     { name = "sentence-transformers", specifier = ">=5.5.0" },
     { name = "setfit", specifier = ">=1.1.3" },
+    { name = "timm", specifier = ">=1.0.27" },
     { name = "torch", marker = "sys_platform != 'linux'", specifier = ">=2.12.0" },
     { name = "torch", marker = "sys_platform == 'linux'", specifier = ">=2.12.0", index = "https://download.pytorch.org/whl/cpu" },
     { name = "transformers", specifier = ">=4.50.0,<6.0.0" },
@@ -856,6 +859,7 @@ detectors = [
     { name = "setfit", specifier = ">=1.1.3" },
     { name = "spacy", specifier = ">=3.8.14" },
     { name = "textstat", specifier = ">=0.7.13" },
+    { name = "timm", specifier = ">=1.0.27" },
     { name = "torch", marker = "sys_platform != 'linux'", specifier = ">=2.12.0" },
     { name = "torch", marker = "sys_platform == 'linux'", specifier = ">=2.12.0", index = "https://download.pytorch.org/whl/cpu" },
     { name = "transformers", specifier = ">=4.50.0,<6.0.0" },
@@ -942,7 +946,7 @@ threat-ml = [
 
 [[package]]
 name = "classifyre-schemas"
-version = "0.4.15"
+version = "0.4.16"
 source = { editable = "../../packages/schemas" }
 dependencies = [
     { name = "fastjsonschema" },
@@ -5349,6 +5353,23 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/16/5b/f2aa703a4fc5d2dff73460a7d46cc2f3f44aa0f3dd8eeb20d2a0ecf68862/tiktoken-0.13.0-cp314-cp314t-win_amd64.whl", hash = "sha256:85b78cc3a2c3d48723ca751fa981f1fedccd54194ca0471b957364353a898b07", size = 918110, upload-time = "2026-05-15T04:51:17.237Z" },
 ]
 
+[[package]]
+name = "timm"
+version = "1.0.27"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "huggingface-hub" },
+    { name = "pyyaml" },
+    { name = "safetensors" },
+    { name = "torch", version = "2.12.0", source = { registry = "https://pypi.org/simple" }, marker = "sys_platform != 'linux'" },
+    { name = "torch", version = "2.12.0+cpu", source = { registry = "https://download.pytorch.org/whl/cpu" }, marker = "sys_platform == 'linux'" },
+    { name = "torchvision" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/08/54/ece85b0eef3700c90db8271a43669b05a0ebbe2edb1962329c34374a297e/timm-1.0.27.tar.gz", hash = "sha256:315dfe63186ca9fb7ff941268941231fd5be259f2b4bb4afa28560ae1015cb9a", size = 2439861, upload-time = "2026-05-08T19:38:36.844Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/1f/2e/26bab7686ff4aed48f8f5f6c23e2aa37b7a37ddd9effe3aa61e908fd518f/timm-1.0.27-py3-none-any.whl", hash = "sha256:5ff07c9ddf53cbada88eab1c93ff175c64cab683b5a2fddf863bcee985926f89", size = 2589280, upload-time = "2026-05-08T19:38:35.034Z" },
+]
+
 [[package]]
 name = "tldextract"
 version = "5.3.1"