feat: support auditing

aperag/api/components/schemas/audit.yaml

@@ -0,0 +1,89 @@
+auditLog:
+  type: object
+  description: Audit log entry
+  properties:
+    id:
+      type: string
+      description: Audit log ID
+    user_id:
+      type: string
+      nullable: true
+      description: User ID who performed the action
+    username:
+      type: string
+      nullable: true
+      description: Username for display
+    resource_type:
+      type: string
+      enum: [collection, document, bot, chat, message, api_key, llm_provider, llm_provider_model, model_service_provider, user, flow, search_test]
+      nullable: true
+      description: Type of resource
+    resource_id:
+      type: string
+      nullable: true
+      description: ID of the resource (extracted at query time)
+    api_name:
+      type: string
+      description: API operation name
+    http_method:
+      type: string
+      description: HTTP method (POST, PUT, DELETE)
+    path:
+      type: string
+      description: API path
+    status_code:
+      type: integer
+      nullable: true
+      description: HTTP status code
+    start_time:
+      type: integer
+      format: int64
+      description: Request start time (milliseconds since epoch)
+    end_time:
+      type: integer
+      format: int64
+      nullable: true
+      description: Request end time (milliseconds since epoch)
+    duration_ms:
+      type: integer
+      nullable: true
+      description: Request duration in milliseconds (calculated)
+    request_data:
+      type: string
+      nullable: true
+      description: Request data (JSON string)
+    response_data:
+      type: string
+      nullable: true
+      description: Response data (JSON string)
+    error_message:
+      type: string
+      nullable: true
+      description: Error message if failed
+    ip_address:
+      type: string
+      nullable: true
+      description: Client IP address
+    user_agent:
+      type: string
+      nullable: true
+      description: User agent string
+    request_id:
+      type: string
+      description: Request ID for tracking
+    created:
+      type: string
+      format: date-time
+      description: Created timestamp
+
+auditLogList:
+  type: object
+  description: List of audit logs
+  properties:
+    items:
+      type: array
+      description: Audit log entries
+      items:
+        $ref: '#/auditLog'
+
+

aperag/api/openapi.yaml

@@ -88,6 +88,12 @@ paths:
   /prompt-templates:
     $ref: './paths/prompt_templates.yaml#/promptTemplates'
 
+  # audit
+  /audit-logs:
+    $ref: './paths/audit.yaml#/audit_logs'
+  /audit-logs/{audit_id}:
+    $ref: './paths/audit.yaml#/audit_log_detail'
+
   # users
   /invite:
     $ref: './paths/auth.yaml#/invite'

aperag/api/paths/audit.yaml

@@ -0,0 +1,73 @@
+audit_logs:
+  get:
+    tags:
+      - audit
+    summary: List audit logs
+    description: List audit logs with filtering options
+    operationId: list_audit_logs
+    parameters:
+      - name: api_name
+        in: query
+        required: false
+        schema:
+          type: string
+        description: Filter by API name
+      - name: start_date
+        in: query
+        required: false
+        schema:
+          type: string
+          format: date-time
+        description: Filter by start date
+      - name: end_date
+        in: query
+        required: false
+        schema:
+          type: string
+          format: date-time
+        description: Filter by end date
+      - name: limit
+        in: query
+        required: false
+        schema:
+          type: integer
+          maximum: 5000
+          default: 1000
+        description: Maximum number of records
+    responses:
+      "200":
+        description: Audit logs retrieved successfully
+        content:
+          application/json:
+            schema:
+              $ref: "../components/schemas/audit.yaml#/auditLogList"
+      "403":
+        description: Admin access required
+
+audit_log_detail:
+  get:
+    tags:
+      - audit
+    summary: Get audit log detail
+    description: Get a specific audit log by ID
+    operationId: get_audit_log
+    parameters:
+      - name: audit_id
+        in: path
+        required: true
+        schema:
+          type: string
+        description: Audit log ID
+    responses:
+      "200":
+        description: Audit log retrieved successfully
+        content:
+          application/json:
+            schema:
+              $ref: "../components/schemas/audit.yaml#/auditLog"
+      "403":
+        description: Admin access required
+      "404":
+        description: Audit log not found
+
+

aperag/app.py

@@ -17,6 +17,7 @@
 from aperag.exception_handlers import register_exception_handlers
 from aperag.llm.litellm_track import register_opik_llm_track
 from aperag.views.api_key import router as api_key_router
+from aperag.views.audit import router as audit_router
 from aperag.views.auth import router as auth_router
 from aperag.views.chat_completion import router as chat_completion_router
 from aperag.views.config import router as config_router
@@ -38,6 +39,7 @@
 app.include_router(auth_router, prefix="/api/v1")
 app.include_router(main_router, prefix="/api/v1")
 app.include_router(api_key_router, prefix="/api/v1")
+app.include_router(audit_router, prefix="/api/v1")  # Add audit router
 app.include_router(flow_router, prefix="/api/v1")
 app.include_router(llm_router, prefix="/api/v1")
 app.include_router(chat_completion_router, prefix="/v1")

aperag/db/models.py

@@ -24,6 +24,7 @@
     Boolean,
     Column,
     DateTime,
+    Index,
     Integer,
     String,
     Text,
@@ -746,3 +747,67 @@ def update_spec(self, desired_state: IndexDesiredState = None, created_by: str =
             self.created_by = created_by
         self.version += 1
         self.gmt_updated = utc_now()
+
+
+class AuditResource(str, Enum):
+    """Audit resource types"""
+
+    COLLECTION = "collection"
+    DOCUMENT = "document"
+    BOT = "bot"
+    CHAT = "chat"
+    MESSAGE = "message"
+    API_KEY = "api_key"
+    LLM_PROVIDER = "llm_provider"
+    LLM_PROVIDER_MODEL = "llm_provider_model"
+    MODEL_SERVICE_PROVIDER = "model_service_provider"
+    USER = "user"
+    CONFIG = "config"
+    INVITATION = "invitation"
+    AUTH = "auth"
+    CHAT_COMPLETION = "chat_completion"
+    SEARCH_TEST = "search_test"
+    LLM = "llm"
+    FLOW = "flow"
+    SYSTEM = "system"
+
+
+class AuditLog(Base):
+    """Audit log model to track all system operations"""
+
+    __tablename__ = "audit_log"
+
+    id = Column(String(36), primary_key=True, default=lambda: str(uuid.uuid4()))
+    user_id = Column(String(36), nullable=True, comment="User ID")
+    username = Column(String(255), nullable=True, comment="Username")
+    resource_type = Column(EnumColumn(AuditResource), nullable=True, comment="Resource type")
+    resource_id = Column(String(255), nullable=True, comment="Resource ID (extracted at query time)")
+    api_name = Column(String(255), nullable=False, comment="API operation name")
+    http_method = Column(String(10), nullable=False, comment="HTTP method (POST, PUT, DELETE)")
+    path = Column(String(512), nullable=False, comment="API path")
+    status_code = Column(Integer, nullable=True, comment="HTTP status code")
+    request_data = Column(Text, nullable=True, comment="Request data (JSON)")
+    response_data = Column(Text, nullable=True, comment="Response data (JSON)")
+    error_message = Column(Text, nullable=True, comment="Error message if failed")
+    ip_address = Column(String(45), nullable=True, comment="Client IP address")
+    user_agent = Column(String(500), nullable=True, comment="User agent string")
+    request_id = Column(String(255), nullable=False, comment="Request ID for tracking")
+    start_time = Column(BigInteger, nullable=False, comment="Request start time (milliseconds since epoch)")
+    end_time = Column(BigInteger, nullable=True, comment="Request end time (milliseconds since epoch)")
+    gmt_created = Column(DateTime(timezone=True), nullable=False, default=utc_now, comment="Created time")
+
+    # Index for better query performance
+    __table_args__ = (
+        Index("idx_audit_user_id", "user_id"),
+        Index("idx_audit_resource_type", "resource_type"),
+        Index("idx_audit_api_name", "api_name"),
+        Index("idx_audit_http_method", "http_method"),
+        Index("idx_audit_status_code", "status_code"),
+        Index("idx_audit_gmt_created", "gmt_created"),
+        Index("idx_audit_resource_id", "resource_id"),
+        Index("idx_audit_request_id", "request_id"),
+        Index("idx_audit_start_time", "start_time"),
+    )
+
+    def __repr__(self):
+        return f"<AuditLog(id={self.id}, user={self.username}, api={self.api_name}, method={self.http_method}, status={self.status_code})>"

aperag/exceptions.py

@@ -12,7 +12,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import traceback
 from enum import Enum
 from http import HTTPStatus
 from typing import Any, Dict, Optional
@@ -90,7 +89,6 @@ def __init__(self, error_code: ErrorCode, message: Optional[str] = None, details
         self.error_code = error_code
         self.message = message or error_code.error_name.replace("_", " ").title()
         self.details = details or {}
-        traceback.print_stack()
         super().__init__(self.message)
 
     @property

aperag/migration/versions/20250621002836-2768dfee8bbc.py

@@ -0,0 +1,70 @@
+"""empty message
+
+Revision ID: 2768dfee8bbc
+Revises: 12ea6d2bf365
+Create Date: 2025-06-21 00:28:36.443046
+
+"""
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision: str = '2768dfee8bbc'
+down_revision: Union[str, None] = '12ea6d2bf365'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    """Upgrade schema."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.create_table('audit_log',
+    sa.Column('id', sa.String(length=36), nullable=False),
+    sa.Column('user_id', sa.String(length=36), nullable=True, comment='User ID'),
+    sa.Column('username', sa.String(length=255), nullable=True, comment='Username'),
+    sa.Column('resource_type', sa.Enum('collection', 'document', 'bot', 'chat', 'message', 'api_key', 'llm_provider', 'llm_provider_model', 'model_service_provider', 'user', 'config', 'invitation', 'auth', 'chat_completion', 'search_test', 'llm', 'flow', 'system', name='auditresource'), nullable=True, comment='Resource type'),
+    sa.Column('resource_id', sa.String(length=255), nullable=True, comment='Resource ID (extracted at query time)'),
+    sa.Column('api_name', sa.String(length=255), nullable=False, comment='API operation name'),
+    sa.Column('http_method', sa.String(length=10), nullable=False, comment='HTTP method (POST, PUT, DELETE)'),
+    sa.Column('path', sa.String(length=512), nullable=False, comment='API path'),
+    sa.Column('status_code', sa.Integer(), nullable=True, comment='HTTP status code'),
+    sa.Column('request_data', sa.Text(), nullable=True, comment='Request data (JSON)'),
+    sa.Column('response_data', sa.Text(), nullable=True, comment='Response data (JSON)'),
+    sa.Column('error_message', sa.Text(), nullable=True, comment='Error message if failed'),
+    sa.Column('ip_address', sa.String(length=45), nullable=True, comment='Client IP address'),
+    sa.Column('user_agent', sa.String(length=500), nullable=True, comment='User agent string'),
+    sa.Column('request_id', sa.String(length=255), nullable=False, comment='Request ID for tracking'),
+    sa.Column('start_time', sa.BigInteger(), nullable=False, comment='Request start time (milliseconds since epoch)'),
+    sa.Column('end_time', sa.BigInteger(), nullable=True, comment='Request end time (milliseconds since epoch)'),
+    sa.Column('gmt_created', sa.DateTime(timezone=True), nullable=False, comment='Created time'),
+    sa.PrimaryKeyConstraint('id')
+    )
+    op.create_index('idx_audit_api_name', 'audit_log', ['api_name'], unique=False)
+    op.create_index('idx_audit_gmt_created', 'audit_log', ['gmt_created'], unique=False)
+    op.create_index('idx_audit_http_method', 'audit_log', ['http_method'], unique=False)
+    op.create_index('idx_audit_request_id', 'audit_log', ['request_id'], unique=False)
+    op.create_index('idx_audit_resource_id', 'audit_log', ['resource_id'], unique=False)
+    op.create_index('idx_audit_resource_type', 'audit_log', ['resource_type'], unique=False)
+    op.create_index('idx_audit_start_time', 'audit_log', ['start_time'], unique=False)
+    op.create_index('idx_audit_status_code', 'audit_log', ['status_code'], unique=False)
+    op.create_index('idx_audit_user_id', 'audit_log', ['user_id'], unique=False)
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    """Downgrade schema."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_index('idx_audit_user_id', table_name='audit_log')
+    op.drop_index('idx_audit_status_code', table_name='audit_log')
+    op.drop_index('idx_audit_start_time', table_name='audit_log')
+    op.drop_index('idx_audit_resource_type', table_name='audit_log')
+    op.drop_index('idx_audit_resource_id', table_name='audit_log')
+    op.drop_index('idx_audit_request_id', table_name='audit_log')
+    op.drop_index('idx_audit_http_method', table_name='audit_log')
+    op.drop_index('idx_audit_gmt_created', table_name='audit_log')
+    op.drop_index('idx_audit_api_name', table_name='audit_log')
+    op.drop_table('audit_log')
+    # ### end Alembic commands ###