microsoft · dworthen · Jul 17, 2026 · Jul 15, 2026 · Jul 15, 2026 · Jul 15, 2026
@@ -1,15 +1,45 @@
-# To get started with Dependabot version updates, you'll need to specify which
-# package ecosystems to update and where the package manifests are located.
+# Dependabot configuration for the GraphRAG uv workspace monorepo.
 # Please see the documentation for all configuration options:
 # https://docs.github.com/code-security/dependabot/dependabot-version-updates/configuration-options-for-the-dependabot.yml-file
 version: 2
 updates:
-  - package-ecosystem: "pip" # See documentation for possible values
-    directory: "/" # Location of package manifests
+  # Python dependencies managed by uv. The workspace root (/) holds the shared
+  # uv.lock and the dev dependency group; each workspace member manifest lives
+  # under packages/*. The unified-search-app project is intentionally excluded
+  # (it lives at the repo root, not under packages/, and is maintained separately).
+  - package-ecosystem: "uv"
+    directories:
+      - "/"
+      - "/packages/*"
     schedule:
       interval: "weekly"
+    open-pull-requests-limit: 10
+    groups:
+      # Azure SDKs tend to move together; keep them in a single PR.
+      azure:
+        patterns:
+          - "azure-*"
+      # Development and release tooling (lint, type-check, test, docs, release).
+      dev-tooling:
+        patterns:
+          - "pytest*"
+          - "ruff"
+          - "pyright"
+          - "ipykernel"
+          - "jupyter"
+          - "nbconvert"
+          - "coverage"
+          - "deptry"
+          - "poethepoet"
+          - "semversioner"
+          - "pandas-stubs"
+          - "mkdocs*"
   - package-ecosystem: "github-actions"
-    # Workflow files stored in the default location of `.github/workflows`. (You don't need to specify `/.github/workflows` for `directory`. You can use `directory: "/"`.)
+    # Workflow files stored in the default location of `.github/workflows`.
     directory: "/"
     schedule:
       interval: "weekly"
+    groups:
+      github-actions:
+        patterns:
+          - "*"
@@ -0,0 +1,4 @@
+{
+  "type": "patch",
+  "description": "Bump dependencies."
+}
@@ -32,7 +32,7 @@ classifiers = [
 dependencies = [
     "python-dotenv~=1.0",
     "pyyaml~=6.0",
-    "toml",
+    "toml~=0.10",
 ]
 
 [project.urls]

@@ -31,10 +31,9 @@ classifiers = [
 ]
 dependencies = [
     "graphrag-common==3.1.0",
-    "graphrag-storage==3.1.0    ",
+    "graphrag-storage==3.1.0",
     "pydantic~=2.10",
-    "markitdown~=0.1.0",
-    "markitdown[pdf]",
+    "markitdown[pdf]~=0.1.0",
     "pyarrow>=14.0.0"
 ]
 

@@ -209,12 +209,12 @@ async def completion_async(
 
     @property
     def metrics_store(self) -> "MetricsStore":
-        """Get metrics store."""
+        """The metrics store."""
         return self._metrics_store
 
     @property
     def tokenizer(self) -> "Tokenizer":
-        """Get tokenizer."""
+        """The tokenizer."""
         return self._tokenizer
 
 

@@ -121,10 +121,10 @@ async def completion_async(
 
     @property
     def metrics_store(self) -> "MetricsStore":
-        """Get metrics store."""
+        """The metrics store."""
         return self._metrics_store
 
     @property
     def tokenizer(self) -> "Tokenizer":
-        """Get tokenizer."""
+        """The tokenizer."""
         return self._tokenizer
@@ -148,12 +148,12 @@ async def embedding_async(
 
     @property
     def metrics_store(self) -> "MetricsStore":
-        """Get metrics store."""
+        """The metrics store."""
         return self._metrics_store
 
     @property
     def tokenizer(self) -> "Tokenizer":
-        """Get tokenizer."""
+        """The tokenizer."""
         return self._tokenizer
 
 

@@ -72,10 +72,10 @@ async def embedding_async(
 
     @property
     def metrics_store(self) -> "MetricsStore":
-        """Get metrics store."""
+        """The metrics store."""
         return self._metrics_store
 
     @property
     def tokenizer(self) -> "Tokenizer":
-        """Get tokenizer."""
+        """The tokenizer."""
         return self._tokenizer
@@ -74,7 +74,7 @@ def _on_exit_(self) -> None:
 
     @property
     def id(self) -> str:
-        """Get the ID of the metrics store."""
+        """The ID of the metrics store."""
         return self._id
 
     def update_metrics(self, *, metrics: "Metrics") -> None:

@@ -41,7 +41,7 @@ def __init__(
     @property
     @abstractmethod
     def id(self) -> str:
-        """Get the ID of the metrics store."""
+        """The ID of the metrics store."""
         raise NotImplementedError
 
     @abstractmethod

@@ -20,7 +20,7 @@ def __init__(
 
     @property
     def id(self) -> str:
-        """Get the ID of the metrics store."""
+        """The ID of the metrics store."""
         return ""
 
     def update_metrics(self, *, metrics: Metrics) -> None:

@@ -100,7 +100,7 @@ class LLMCompletionResponse(ChatCompletion, Generic[ResponseFormat]):
     @computed_field
     @property
     def content(self) -> str:
-        """Get the content of the first choice message."""
+        """The content of the first choice message."""
         return self.choices[0].message.content or ""
 
 
@@ -191,13 +191,13 @@ class LLMEmbeddingResponse(CreateEmbeddingResponse):
     @computed_field
     @property
     def embeddings(self) -> list[list[float]]:
-        """Get the embeddings as a list of lists of floats."""
+        """The embeddings as a list of lists of floats."""
         return [data.embedding for data in self.data]
 
     @computed_field
     @property
     def first_embedding(self) -> list[float]:
-        """Get the first embedding."""
+        """The first embedding."""
         return self.embeddings[0] if self.embeddings else []
 
 

@@ -173,7 +173,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": null,
    "id": "21e0e1e4",
    "metadata": {},
    "outputs": [],
@@ -203,7 +203,7 @@
     "    messages=\"It is sunny and 52 degrees fahrenheit in Seattle. It is cloudy and 75 degrees fahrenheit in San Francisco.\",\n",
     "    response_format=WeatherReports,\n",
     ")  # type: ignore\n",
-    "response: LLMCompletionResponse[WeatherReports] = llm_completion.completion(  # type: ignore\n",
+    "response: LLMCompletionResponse[WeatherReports] = llm_completion.completion(  # type: ignore  # noqa: F811\n",
     "    messages=\"It is sunny and 52 degrees fahrenheit in Seattle. It is cloudy and 75 degrees fahrenheit in San Francisco.\",\n",
     "    response_format=WeatherReports,\n",
     ")  # type: ignore\n",
@@ -228,7 +228,7 @@
     "\n",
     "llm_completion.metrics_store.clear_metrics()\n",
     "# Same request but different response format. Should not hit cache.\n",
-    "response: LLMCompletionResponse[WeatherReports2] = llm_completion.completion(\n",
+    "response: LLMCompletionResponse[WeatherReports2] = llm_completion.completion(  # noqa: F811\n",
     "    messages=\"It is sunny and 52 degrees fahrenheit in Seattle. It is cloudy and 75 degrees fahrenheit in San Francisco.\",\n",
     "    response_format=WeatherReports2,\n",
     ")  # type: ignore\n",

@@ -33,13 +33,17 @@ classifiers = [
 ]
 dependencies = [
     "azure-identity~=1.25",
+    # fastapi is required for litellm tool calling. https://github.com/BerriAI/litellm/issues/32993
+    "fastapi>=0.136.3,<1.0",
     "graphrag-cache==3.1.0",
     "graphrag-common==3.1.0",
     "jinja2~=3.1",
-    "litellm==1.86.2",
+    "litellm==1.92.0",
     "nest-asyncio2~=1.7",
+    # orjson is required for litellm tool calling. https://github.com/BerriAI/litellm/issues/32993
+    "orjson>=3.11.6,<4.0",
     "pydantic~=2.10",
-    "typing-extensions~=4.12"
+    "typing-extensions~=4.12",
 ]
 
 [project.urls]

@@ -102,10 +102,13 @@ async def _aiter_impl(self) -> AsyncIterator[Any]:
         """Implement async iteration over rows."""
         if isinstance(self._storage, FileStorage):
             file_path = self._storage.get_path(self._file_key)
-            with Path.open(file_path, "r", encoding=self._encoding) as f:
+            f = Path.open(file_path, "r", encoding=self._encoding)
+            try:
                 reader = csv.DictReader(f)
                 for row in reader:
                     yield _apply_transformer(self._transformer, row)
+            finally:
+                f.close()
 
     async def length(self) -> int:
         """Return the number of rows in the table."""

@@ -30,12 +30,12 @@ classifiers = [
     "Programming Language :: Python :: 3.13",
 ]
 dependencies = [
-    "aiofiles~=24.1",
+    "aiofiles~=25.1",
     "azure-cosmos~=4.9",
     "azure-identity~=1.25",
     "azure-storage-blob~=12.24",
     "graphrag-common==3.1.0",
-    "pandas~=2.3",
+    "pandas~=3.0",
     "pydantic~=2.10",
 ]
 

@@ -33,11 +33,11 @@ dependencies = [
     "azure-core~=1.32",
     "azure-cosmos~=4.9",
     "azure-identity~=1.25",
-    "azure-search-documents~=11.6",
+    "azure-search-documents~=12.0",
     "graphrag-common==3.1.0",
-    "lancedb~=0.24.1",
+    "lancedb~=0.34.0",
     "numpy~=2.1",
-    "pyarrow~=22.0",
+    "pyarrow~=25.0",
     "pydantic~=2.10",
 ]
 

@@ -140,7 +140,7 @@ def _modularity_components(
         degree_sums_for[tgt_comm] += weight
         total_edge_weight += weight
 
-    if total_edge_weight == 0.0:
+    if total_edge_weight <= 0.0:
         return dict.fromkeys(communities, 0.0)
 
     return {

@@ -19,7 +19,6 @@ def finalize_community_reports(
         communities.loc[:, ["community", "parent", "children", "size", "period"]],
         on="community",
         how="left",
-        copy=False,
     )
 
     community_reports["community"] = community_reports["community"].astype(int)

@@ -51,7 +51,7 @@ def __exit__(
 
     @property
     def metrics(self) -> WorkflowMetrics:
-        """Return collected metrics as a WorkflowMetrics dataclass."""
+        """The collected metrics as a WorkflowMetrics dataclass."""
         return WorkflowMetrics(
             overall=self._elapsed,
             peak_memory_bytes=self._peak_memory,

@@ -48,7 +48,7 @@ def _update_and_merge_communities(
     # Increment only the non-NaN values in delta_communities["community"]
     community_id_mapping = {
         v: v + old_max_community_id + 1
-        for k, v in delta_communities["community"].dropna().astype(int).items()
+        for v in delta_communities["community"].dropna().astype(int)
     }
     community_id_mapping.update({-1: -1})
 
@@ -69,7 +69,7 @@ def _update_and_merge_communities(
 
     # Merge the final communities
     merged_communities = pd.concat(
-        [old_communities, delta_communities], ignore_index=True, copy=False
+        [old_communities, delta_communities], ignore_index=True
     )
 
     # Rename title
@@ -136,7 +136,7 @@ def _update_and_merge_community_reports(
 
     # Merge the final community reports
     merged_community_reports = pd.concat(
-        [old_community_reports, delta_community_reports], ignore_index=True, copy=False
+        [old_community_reports, delta_community_reports], ignore_index=True
     )
 
     # Maintain type compat with query

@@ -35,7 +35,6 @@ def _group_and_resolve_entities(
         old_entities_df[["id", "title"]],
         on="title",
         suffixes=("_B", "_A"),
-        copy=False,
     )
     id_mapping = dict(zip(merged["id_B"], merged["id_A"], strict=True))
 
@@ -45,9 +44,7 @@ def _group_and_resolve_entities(
         initial_id, initial_id + len(delta_entities_df)
     )
     # Concat A and B
-    combined = pd.concat(
-        [old_entities_df, delta_entities_df], ignore_index=True, copy=False
-    )
+    combined = pd.concat([old_entities_df, delta_entities_df], ignore_index=True)
 
     # Group by title and resolve conflicts
     aggregated = (

@@ -71,7 +71,7 @@ async def concat_dataframes(
     # Merge the final documents
     initial_id = old_df["human_readable_id"].max() + 1
     delta_df["human_readable_id"] = np.arange(initial_id, initial_id + len(delta_df))
-    final_df = pd.concat([old_df, delta_df], ignore_index=True, copy=False)
+    final_df = pd.concat([old_df, delta_df], ignore_index=True)
 
     await output_table_provider.write_dataframe(name, final_df)
 

@@ -45,7 +45,7 @@ def _update_and_merge_relationships(
 
     # Merge the DataFrames without copying if possible
     merged_relationships = pd.concat(
-        [old_relationships, delta_relationships], ignore_index=True, copy=False
+        [old_relationships, delta_relationships], ignore_index=True
     )
 
     # Group by title and resolve conflicts

@@ -77,4 +77,4 @@ def _merge_covariates(
     )
 
     # Concatenate the old and delta covariates
-    return pd.concat([old_covariates, delta_covariates], ignore_index=True, copy=False)
+    return pd.concat([old_covariates, delta_covariates], ignore_index=True)
@@ -92,4 +92,4 @@ def _update_and_merge_text_units(
         initial_id, initial_id + len(delta_text_units)
     )
     # Merge the final text units
-    return pd.concat([old_text_units, delta_text_units], ignore_index=True, copy=False)
+    return pd.concat([old_text_units, delta_text_units], ignore_index=True)
@@ -113,7 +113,9 @@ def _write_log(self, log: dict[str, Any]):
         blob_client = self._blob_service_client.get_blob_client(
             self._container_name, self._blob_name
         )
-        blob_client.append_block(json.dumps(log, indent=4, ensure_ascii=False) + "\n")
+        blob_client.append_block(
+            (json.dumps(log, indent=4, ensure_ascii=False) + "\n").encode("utf-8")
+        )
 
         # update the blob's block count
         self._num_blocks += 1