Fix exception handling across cognee - lost exception stacktraces and exception metadata (#1518)

<!-- .github/pull_request_template.md -->

## Description
<!--
Please provide a clear, human-generated description of the changes in
this PR.
DO NOT use AI-generated descriptions. We want to understand your thought
process and reasoning.
-->

In several places, our exception handling discards valuable debugging
context - either by losing the original stack trace or omitting metadata
from the underlying error. This makes diagnosing root causes
significantly harder.

This PR updates exception throwing to use [`raise ...
from`](https://docs.python.org/3/reference/simple_stmts.html#raise),
which:
1. The original stack trace is always preserved, even when wrapping
exceptions
2. Custom exceptions explicitly reference their underlying cause, making
error chains clearer



## Type of Change
<!-- Please check the relevant option -->
- [ ] Bug fix (non-breaking change that fixes an issue)
- [ ] New feature (non-breaking change that adds functionality)
- [ ] Breaking change (fix or feature that would cause existing
functionality to change)
- [ ] Documentation update
- [ ] Code refactoring
- [ ] Performance improvement
- [ ] Other (please specify):

## Screenshots/Videos (if applicable)
<!-- Add screenshots or videos to help explain your changes -->

## Pre-submission Checklist
<!-- Please check all boxes that apply before submitting your PR -->
- [ ] **I have tested my changes thoroughly before submitting this PR**
- [ ] **This PR contains minimal changes necessary to address the
issue/feature**
- [ ] My code follows the project's coding standards and style
guidelines
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] I have added necessary documentation (if applicable)
- [ ] All new and existing tests pass
- [ ] I have searched existing PRs to ensure this change hasn't been
submitted already
- [ ] I have linked any relevant issues in the description
- [ ] My commits have clear and descriptive messages

## DCO Affirmation
I affirm that all code in every commit of this pull request conforms to
the terms of the Topoteretes Developer Certificate of Origin.

cognee/cli/commands/add_command.py

@@ -70,11 +70,11 @@ After adding data, use `cognee cognify` to process it into knowledge graphs.
                     await cognee.add(data=data_to_add, dataset_name=args.dataset_name)
                     fmt.success(f"Successfully added data to dataset '{args.dataset_name}'")
                 except Exception as e:
-                    raise CliCommandInnerException(f"Failed to add data: {str(e)}")
+                    raise CliCommandInnerException(f"Failed to add data: {str(e)}") from e
 
             asyncio.run(run_add())
 
         except Exception as e:
             if isinstance(e, CliCommandInnerException):
-                raise CliCommandException(str(e), error_code=1)
+                raise CliCommandException(str(e), error_code=1) from e
-            raise CliCommandException(f"Error adding data: {str(e)}", error_code=1)
+            raise CliCommandException(f"Error adding data: {str(e)}", error_code=1) from e

cognee/cli/commands/cognify_command.py

@@ -107,7 +107,7 @@ After successful cognify processing, use `cognee search` to query the knowledge
                     )
                     return result
                 except Exception as e:
-                    raise CliCommandInnerException(f"Failed to cognify: {str(e)}")
+                    raise CliCommandInnerException(f"Failed to cognify: {str(e)}") from e
 
             result = asyncio.run(run_cognify())
 
@@ -124,5 +124,5 @@ After successful cognify processing, use `cognee search` to query the knowledge
 
         except Exception as e:
             if isinstance(e, CliCommandInnerException):
-                raise CliCommandException(str(e), error_code=1)
+                raise CliCommandException(str(e), error_code=1) from e
-            raise CliCommandException(f"Error during cognification: {str(e)}", error_code=1)
+            raise CliCommandException(f"Error during cognification: {str(e)}", error_code=1) from e

cognee/cli/commands/config_command.py

@@ -79,8 +79,10 @@ Configuration changes will affect how cognee processes and stores data.
 
         except Exception as e:
             if isinstance(e, CliCommandInnerException):
-                raise CliCommandException(str(e), error_code=1)
+                raise CliCommandException(str(e), error_code=1) from e
-            raise CliCommandException(f"Error managing configuration: {str(e)}", error_code=1)
+            raise CliCommandException(
+                f"Error managing configuration: {str(e)}", error_code=1
+            ) from e
 
     def _handle_get(self, args: argparse.Namespace) -> None:
         try:
@@ -122,7 +124,7 @@ Configuration changes will affect how cognee processes and stores data.
                     fmt.note("Configuration viewing not fully implemented yet")
 
         except Exception as e:
-            raise CliCommandInnerException(f"Failed to get configuration: {str(e)}")
+            raise CliCommandInnerException(f"Failed to get configuration: {str(e)}") from e
 
     def _handle_set(self, args: argparse.Namespace) -> None:
         try:
@@ -141,7 +143,7 @@ Configuration changes will affect how cognee processes and stores data.
                 fmt.error(f"Failed to set configuration key '{args.key}'")
 
         except Exception as e:
-            raise CliCommandInnerException(f"Failed to set configuration: {str(e)}")
+            raise CliCommandInnerException(f"Failed to set configuration: {str(e)}") from e
 
     def _handle_unset(self, args: argparse.Namespace) -> None:
         try:
@@ -189,7 +191,7 @@ Configuration changes will affect how cognee processes and stores data.
                 fmt.note("Use 'cognee config list' to see all available configuration options")
 
         except Exception as e:
-            raise CliCommandInnerException(f"Failed to unset configuration: {str(e)}")
+            raise CliCommandInnerException(f"Failed to unset configuration: {str(e)}") from e
 
     def _handle_list(self, args: argparse.Namespace) -> None:
         try:
@@ -209,7 +211,7 @@ Configuration changes will affect how cognee processes and stores data.
             fmt.echo("  cognee config reset         - Reset all to defaults")
 
         except Exception as e:
-            raise CliCommandInnerException(f"Failed to list configuration: {str(e)}")
+            raise CliCommandInnerException(f"Failed to list configuration: {str(e)}") from e
 
     def _handle_reset(self, args: argparse.Namespace) -> None:
         try:
@@ -222,4 +224,4 @@ Configuration changes will affect how cognee processes and stores data.
             fmt.echo("This would reset all settings to their default values")
 
         except Exception as e:
-            raise CliCommandInnerException(f"Failed to reset configuration: {str(e)}")
+            raise CliCommandInnerException(f"Failed to reset configuration: {str(e)}") from e

cognee/cli/commands/delete_command.py

@@ -100,7 +100,7 @@ Be careful with deletion operations as they are irreversible.
                     else:
                         await cognee.delete(dataset_name=args.dataset_name, user_id=args.user_id)
                 except Exception as e:
-                    raise CliCommandInnerException(f"Failed to delete: {str(e)}")
+                    raise CliCommandInnerException(f"Failed to delete: {str(e)}") from e
 
             asyncio.run(run_delete())
             # This success message may be inaccurate due to the underlying bug, but we leave it for now.
@@ -108,5 +108,5 @@ Be careful with deletion operations as they are irreversible.
 
         except Exception as e:
             if isinstance(e, CliCommandInnerException):
-                raise CliCommandException(str(e), error_code=1)
+                raise CliCommandException(str(e), error_code=1) from e
-            raise CliCommandException(f"Error deleting data: {str(e)}", error_code=1)
+            raise CliCommandException(f"Error deleting data: {str(e)}", error_code=1) from e

cognee/cli/commands/search_command.py

@@ -104,7 +104,7 @@ Search Types & Use Cases:
                     )
                     return results
                 except Exception as e:
-                    raise CliCommandInnerException(f"Failed to search: {str(e)}")
+                    raise CliCommandInnerException(f"Failed to search: {str(e)}") from e
 
             results = asyncio.run(run_search())
 
@@ -141,5 +141,5 @@ Search Types & Use Cases:
 
         except Exception as e:
             if isinstance(e, CliCommandInnerException):
-                raise CliCommandException(str(e), error_code=1)
+                raise CliCommandException(str(e), error_code=1) from e
-            raise CliCommandException(f"Error searching: {str(e)}", error_code=1)
+            raise CliCommandException(f"Error searching: {str(e)}", error_code=1) from e

cognee/infrastructure/databases/graph/neptune_driver/neptune_utils.py

@@ -53,7 +53,7 @@ def parse_neptune_url(url: str) -> Tuple[str, str]:
         return graph_id, region
 
     except Exception as e:
-        raise ValueError(f"Failed to parse Neptune Analytics URL '{url}': {str(e)}")
+        raise ValueError(f"Failed to parse Neptune Analytics URL '{url}': {str(e)}") from e
 
 
 def validate_graph_id(graph_id: str) -> bool:

cognee/infrastructure/databases/relational/sqlalchemy/SqlAlchemyAdapter.py

@@ -283,7 +283,7 @@ class SQLAlchemyAdapter:
             try:
                 data_entity = (await session.scalars(select(Data).where(Data.id == data_id))).one()
             except (ValueError, NoResultFound) as e:
-                raise EntityNotFoundError(message=f"Entity not found: {str(e)}")
+                raise EntityNotFoundError(message=f"Entity not found: {str(e)}") from e
 
             # Check if other data objects point to the same raw data location
             raw_data_location_entities = (

cognee/infrastructure/databases/vector/embeddings/FastembedEmbeddingEngine.py

@@ -90,7 +90,9 @@ class FastembedEmbeddingEngine(EmbeddingEngine):
 
         except Exception as error:
             logger.error(f"Embedding error in FastembedEmbeddingEngine: {str(error)}")
-            raise EmbeddingException(f"Failed to index data points using model {self.model}")
+            raise EmbeddingException(
+                f"Failed to index data points using model {self.model}"
+            ) from error
 
     def get_vector_size(self) -> int:
         """

cognee/infrastructure/databases/vector/embeddings/LiteLLMEmbeddingEngine.py

@@ -150,7 +150,7 @@ class LiteLLMEmbeddingEngine(EmbeddingEngine):
             litellm.exceptions.NotFoundError,
         ) as e:
             logger.error(f"Embedding error with model {self.model}: {str(e)}")
-            raise EmbeddingException(f"Failed to index data points using model {self.model}")
+            raise EmbeddingException(f"Failed to index data points using model {self.model}") from e
 
         except Exception as error:
             logger.error("Error embedding text: %s", str(error))

cognee/modules/users/methods/get_authenticated_user.py

@@ -37,6 +37,8 @@ async def get_authenticated_user(
         except Exception as e:
             # Convert any get_default_user failure into a proper HTTP 500 error
             logger.error(f"Failed to create default user: {str(e)}")
-            raise HTTPException(status_code=500, detail=f"Failed to create default user: {str(e)}")
+            raise HTTPException(
+                status_code=500, detail=f"Failed to create default user: {str(e)}"
+            ) from e
 
     return user

cognee/modules/users/roles/methods/create_role.py

@@ -40,8 +40,8 @@ async def create_role(
             # Add association directly to the association table
             role = Role(name=role_name, tenant_id=tenant.id)
             session.add(role)
-        except IntegrityError:
+        except IntegrityError as e:
-            raise EntityAlreadyExistsError(message="Role already exists for tenant.")
+            raise EntityAlreadyExistsError(message="Role already exists for tenant.") from e
 
         await session.commit()
         await session.refresh(role)

cognee/modules/users/tenants/methods/create_tenant.py

@@ -35,5 +35,5 @@ async def create_tenant(tenant_name: str, user_id: UUID) -> UUID:
             await session.merge(user)
             await session.commit()
             return tenant.id
-        except IntegrityError:
+        except IntegrityError as e:
-            raise EntityAlreadyExistsError(message="Tenant already exists.")
+            raise EntityAlreadyExistsError(message="Tenant already exists.") from e