feat(gateway,agents-api): Add scalar for documentation

Signed-off-by: Diwank Tomer <[email protected]>

.env.example

@@ -7,7 +7,6 @@ TEMPORAL_POSTGRES_PASSWORD=<your_temporal_postgres_password>
 LITELLM_POSTGRES_PASSWORD=<your_litellm_postgres_password>
 LITELLM_MASTER_KEY=<your_litellm_master_key>
 LITELLM_REDIS_PASSWORD=<your_litellm_redis_password>
-SKIP_CHECK_DEVELOPER_HEADERS=true
 EMBEDDING_SERVICE_BASE=http://text-embeddings-inference-<gpu|cpu>  # Use the 'gpu' profile to run on GPU
 
 # Memory Store
@@ -16,7 +15,7 @@ EMBEDDING_SERVICE_BASE=http://text-embeddings-inference-<gpu|cpu>  # Use the 'gp
 # COZO_HOST=http://memory-store:9070
 # COZO_PORT=9070
 # COZO_ROCKSDB_DIR=cozo.db
-# COZO_BACKUP_DIR=backup
+# COZO_BACKUP_DIR=/backup
 # COZO_MNT_DIR=/data
 
 # Gateway
@@ -28,6 +27,9 @@ EMBEDDING_SERVICE_BASE=http://text-embeddings-inference-<gpu|cpu>  # Use the 'gp
 # Agents API
 # ---------
 
+# AGENTS_API_MULTI_TENANT_MODE=false
+# AGENTS_API_HOSTNAME=localhost
+# AGENTS_API_PREFIX=""
 # AGENTS_API_KEY_HEADER_NAME=Authorization
 # AGENTS_API_URL=http://agents-api:8080
 # TRUNCATE_EMBED_TEXT=true

agents-api/agents_api/dependencies/developer_id.py

@@ -10,7 +10,7 @@
 
 
 async def get_developer_id(
-    x_developer_id: Annotated[UUID | None, Header()] = None,
+    x_developer_id: Annotated[UUID | None, Header(include_in_schema=False)] = None,
 ) -> UUID:
     if skip_check_developer_headers:
         return x_developer_id or UUID("00000000-0000-0000-0000-000000000000")
@@ -30,7 +30,7 @@ async def get_developer_id(
 
 
 async def get_developer_data(
-    x_developer_id: Annotated[UUID | None, Header()] = None,
+    x_developer_id: Annotated[UUID | None, Header(include_in_schema=False)] = None,
 ) -> Developer:
     if skip_check_developer_headers:
         x_developer_id = x_developer_id or UUID("00000000-0000-0000-0000-000000000000")

agents-api/agents_api/env.py

@@ -12,6 +12,12 @@
 # Initialize the Env object for environment variable parsing.
 env: Any = Env()
 
+# App
+# ---
+multi_tenant_mode: bool = env.bool("AGENTS_API_MULTI_TENANT_MODE", default=False)
+hostname: str = env.str("AGENTS_API_HOSTNAME", default="localhost")
+api_prefix: str = env.str("AGENTS_API_PREFIX", default="")
+
 
 # Debug
 # -----
@@ -39,7 +45,7 @@
 
 api_key_header_name: str = env.str("AGENTS_API_KEY_HEADER_NAME", default="X-Auth-Key")
 skip_check_developer_headers: bool = env.bool(
-    "SKIP_CHECK_DEVELOPER_HEADERS", default=False
+    "SKIP_CHECK_DEVELOPER_HEADERS", default=multi_tenant_mode
 )
 
 
@@ -57,7 +63,7 @@
 embedding_model_id: str = env.str(
     "EMBEDDING_MODEL_ID", default="Alibaba-NLP/gte-large-en-v1.5"
 )
-truncate_embed_text: bool = env.bool("TRUNCATE_EMBED_TEXT", default=False)
+truncate_embed_text: bool = env.bool("TRUNCATE_EMBED_TEXT", default=True)
 
 
 # Temporal
@@ -73,14 +79,17 @@
 # Consolidate environment variables
 environment: Dict[str, Any] = dict(
     debug=debug,
+    multi_tenant_mode=multi_tenant_mode,
+    skip_check_developer_headers=skip_check_developer_headers,
     cozo_host=cozo_host,
     cozo_auth=cozo_auth,
     sentry_dsn=sentry_dsn,
     temporal_endpoint=temporal_endpoint,
     temporal_task_queue=temporal_task_queue,
     api_key=api_key,
     api_key_header_name=api_key_header_name,
-    skip_check_developer_headers=skip_check_developer_headers,
+    hostname=hostname,
+    api_prefix=api_prefix,
     embedding_service_base=embedding_service_base,
     truncate_embed_text=truncate_embed_text,
     temporal_worker_url=temporal_worker_url,
@@ -89,15 +98,15 @@
     testing=testing,
 )
 
-if debug or testing:
-    # Print the loaded environment variables for debugging purposes.
-    print("Environment variables:")
-    pprint(environment)
-    print()
-
-    # Yell if testing is enabled
-    print("@" * 80)
-    print(
-        f"@@@ Running in {'testing' if testing else 'debug'} mode. This should not be enabled in production. @@@"
-    )
-    print("@" * 80)
+# if debug or testing:
+# Print the loaded environment variables for debugging purposes.
+print("Environment variables:")
+pprint(environment)
+print()
+
+# Yell if testing is enabled
+print("@" * 80)
+print(
+    f"@@@ Running in {'testing' if testing else 'debug'} mode. This should not be enabled in production. @@@"
+)
+print("@" * 80)

agents-api/agents_api/web.py

@@ -20,7 +20,7 @@
 
 from .common.exceptions import BaseCommonException
 from .dependencies.auth import get_api_key
-from .env import sentry_dsn
+from .env import api_prefix, hostname, sentry_dsn
 from .exceptions import PromptTooBigError
 from .routers import (
     agents,
@@ -84,22 +84,37 @@ def register_exceptions(app: FastAPI) -> None:
 #       Because some routes don't require auth
 # See: https://fastapi.tiangolo.com/tutorial/bigger-applications/
 #
-app: Any = FastAPI(docs_url="/swagger")
+app: Any = FastAPI(
+    docs_url="/swagger",
+    openapi_prefix=api_prefix,
+    redoc_url=None,
+    title="Julep Agents API",
+    description="API for Julep Agents",
+    version="0.4.0",
+    terms_of_service="https://www.julep.ai/terms",
+    contact={
+        "name": "Julep",
+        "url": "https://www.julep.ai",
+        "email": "[email protected]",
+    },
+    root_path=api_prefix,
+)
 
 # Create a new router for the docs
-docs_router = APIRouter()
+scalar_router = APIRouter()
 
 
-@docs_router.get("/docs", include_in_schema=False)
+@scalar_router.get("/docs", include_in_schema=False)
 async def scalar_html():
     return get_scalar_api_reference(
         openapi_url=app.openapi_url[1:],  # Remove leading '/'
         title=app.title,
+        servers=[{"url": f"http://{hostname}{api_prefix}"}],
     )
 
 
 # Add the docs_router without dependencies
-app.include_router(docs_router)
+app.include_router(scalar_router)
 
 # Add other routers with the get_api_key dependency
 app.include_router(agents.router, dependencies=[Depends(get_api_key)])

agents-api/docker-compose.yml

@@ -15,23 +15,25 @@ x--text-embeddings-inference: &text-embeddings-inference
 # Shared environment variables
 x-shared-environment: &shared-environment
   AGENTS_API_KEY: ${AGENTS_API_KEY}
-  LITELLM_MASTER_KEY: ${LITELLM_MASTER_KEY}
-  COZO_AUTH_TOKEN: ${COZO_AUTH_TOKEN}
-  SKIP_CHECK_DEVELOPER_HEADERS: ${SKIP_CHECK_DEVELOPER_HEADERS:-True}
   AGENTS_API_KEY_HEADER_NAME: ${AGENTS_API_KEY_HEADER_NAME:-Authorization}
+  AGENTS_API_HOSTNAME: ${AGENTS_API_HOSTNAME:-localhost}
+  AGENTS_API_PREFIX: ${AGENTS_API_PREFIX:-}
   AGENTS_API_URL: ${AGENTS_API_URL:-http://agents-api:8080}
-  TRUNCATE_EMBED_TEXT: ${TRUNCATE_EMBED_TEXT:-False}
-  WORKER_URL: ${WORKER_URL:-temporal:7233}
+  COZO_AUTH_TOKEN: ${COZO_AUTH_TOKEN}
+  COZO_HOST: ${COZO_HOST:-http://memory-store:9070}
   DEBUG: ${AGENTS_API_DEBUG:-False}
-  EMBEDDING_SERVICE_BASE: ${EMBEDDING_SERVICE_BASE:-http://text-embeddings-inference}
   EMBEDDING_MODEL_ID: ${EMBEDDING_MODEL_ID:-Alibaba-NLP/gte-large-en-v1.5}
+  EMBEDDING_SERVICE_BASE: ${EMBEDDING_SERVICE_BASE:-http://text-embeddings-inference}
+  LITELLM_MASTER_KEY: ${LITELLM_MASTER_KEY}
   LITELLM_URL: ${LITELLM_URL:-http://litellm:4000}
-  COZO_HOST: ${COZO_HOST:-http://memory-store:9070}
+  SKIP_CHECK_DEVELOPER_HEADERS: ${SKIP_CHECK_DEVELOPER_HEADERS:-True}
   SUMMARIZATION_MODEL_NAME: ${SUMMARIZATION_MODEL_NAME:-gpt-4-turbo}
-  TEMPORAL_WORKER_URL: ${TEMPORAL_WORKER_URL:-temporal:7233}
-  TEMPORAL_NAMESPACE: ${TEMPORAL_NAMESPACE:-default}
   TEMPORAL_ENDPOINT: ${TEMPORAL_ENDPOINT:-temporal:7233}
+  TEMPORAL_NAMESPACE: ${TEMPORAL_NAMESPACE:-default}
   TEMPORAL_TASK_QUEUE: ${TEMPORAL_TASK_QUEUE:-julep-task-queue}
+  TEMPORAL_WORKER_URL: ${TEMPORAL_WORKER_URL:-temporal:7233}
+  TRUNCATE_EMBED_TEXT: ${TRUNCATE_EMBED_TEXT:-True}
+  WORKER_URL: ${WORKER_URL:-temporal:7233}
 
 services:
   agents-api:
@@ -121,8 +123,7 @@ services:
       dockerfile: Dockerfile.migration
     restart: "no" # Make sure to double quote this
     environment:
-      - COZO_HOST=${COZO_HOST:-http://cozo:9070}
-      - COZO_AUTH_TOKEN=${COZO_AUTH_TOKEN:-myauthkey}
+      <<: *shared-environment
 
     develop:
       watch:

gateway/docker-compose.yml

@@ -9,10 +9,10 @@ services:
       # - MODEL_API_URL=${MODEL_API_URL}
       # - MODEL_API_KEY=${MODEL_API_KEY}
       # - MODEL_API_KEY_HEADER_NAME=${MODEL_API_KEY_HEADER_NAME}
-      - AGENTS_API_URL=${AGENTS_API_URL}
+      - AGENTS_API_URL=${AGENTS_API_URL:-http://agents-api:8080}
       - AGENTS_API_KEY=${AGENTS_API_KEY}
-      - AGENTS_API_KEY_HEADER_NAME=${AGENTS_API_KEY_HEADER_NAME}
-      - TRAEFIK_LOG_LEVEL=${TRAEFIK_LOG_LEVEL}
+      - AGENTS_API_KEY_HEADER_NAME=${AGENTS_API_KEY_HEADER_NAME:-Authorization}
+      - TRAEFIK_LOG_LEVEL=${TRAEFIK_LOG_LEVEL:-INFO}
     container_name: gateway
     depends_on:
       agents-api:

gateway/traefik.yml.template

@@ -32,7 +32,7 @@ http:
     agents-api-docs:
       entryPoints:
         - web
-      rule: Path(`/api/{path:docs|openapi.json}`)
+      rule: Path(`/api/{path:swagger|redoc|docs|scalar|openapi.json}`)
       middlewares:
         - agents-api-strip-prefix-api
       service: service-agents-api

llm-proxy/docker-compose.yml

@@ -10,13 +10,13 @@ services:
       - "4000:4000"
     environment:
       - LITELLM_MASTER_KEY=${LITELLM_MASTER_KEY}
-      - LITELLM_POSTGRES_DB=${LITELLM_POSTGRES_DB}
-      - LITELLM_POSTGRES_USER=${LITELLM_POSTGRES_USER}
+      - LITELLM_POSTGRES_DB=${LITELLM_POSTGRES_DB:-litellm}
+      - LITELLM_POSTGRES_USER=${LITELLM_POSTGRES_USER:-llmproxy}
       - LITELLM_POSTGRES_PASSWORD=${LITELLM_POSTGRES_PASSWORD}
-      - LITELLM_REDIS_HOST=${LITELLM_REDIS_HOST}
-      - LITELLM_REDIS_PORT=${LITELLM_REDIS_PORT}
+      - LITELLM_REDIS_HOST=${LITELLM_REDIS_HOST:-litellm-redis}
+      - LITELLM_REDIS_PORT=${LITELLM_REDIS_PORT:-6379}
       - LITELLM_REDIS_PASSWORD=${LITELLM_REDIS_PASSWORD}
-      - LITELLM_DATABASE_URL=postgresql://${LITELLM_POSTGRES_USER}:${LITELLM_POSTGRES_PASSWORD}@litellm-db:5432/${LITELLM_POSTGRES_DB}
+      - LITELLM_DATABASE_URL=postgresql://${LITELLM_POSTGRES_USER:-llmproxy}:${LITELLM_POSTGRES_PASSWORD}@litellm-db:5432/${LITELLM_POSTGRES_DB:-litellm}
       - OPENAI_API_KEY=${OPENAI_API_KEY}
       - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}
       - GROQ_API_KEY=${GROQ_API_KEY}
@@ -49,11 +49,12 @@ services:
     ports:
       - "15432:5432"
     environment:
-      - POSTGRES_DB=${LITELLM_POSTGRES_DB}
-      - POSTGRES_USER=${LITELLM_POSTGRES_USER}
+      - POSTGRES_DB=${LITELLM_POSTGRES_DB:-litellm}
+      - POSTGRES_USER=${LITELLM_POSTGRES_USER:-llmproxy}
       - POSTGRES_PASSWORD=${LITELLM_POSTGRES_PASSWORD}
     healthcheck:
-      test: [ "CMD-SHELL", "pg_isready -d ${LITELLM_POSTGRES_DB} -U ${LITELLM_POSTGRES_USER}" ]
+      test: [ "CMD-SHELL", "pg_isready -d ${LITELLM_POSTGRES_DB:-litellm} -U ${LITELLM_POSTGRES_USER:-llmproxy}" ]
+
       interval: 1s
       timeout: 5s
       retries: 10

scheduler/docker-compose.yml

@@ -8,7 +8,7 @@ services:
       - DB=postgres12
       - DB_PORT=5432
       - DB_HOST=temporal-db
-      - POSTGRES_USER=${TEMPORAL_POSTGRES_USER}
+      - POSTGRES_USER=${TEMPORAL_POSTGRES_USER:-temporal}
       - POSTGRES_PWD=${TEMPORAL_POSTGRES_PASSWORD}
       - POSTGRES_SEEDS=temporal-db
       - DYNAMIC_CONFIG_FILE_PATH=config/dynamicconfig/development-sql.yaml
@@ -31,15 +31,16 @@ services:
   temporal-db:
     image: postgres:16
     environment:
-      - POSTGRES_DB=${TEMPORAL_POSTGRES_DB}
-      - POSTGRES_USER=${TEMPORAL_POSTGRES_USER}
+      - POSTGRES_DB=${TEMPORAL_POSTGRES_DB:-temporal}
+      - POSTGRES_USER=${TEMPORAL_POSTGRES_USER:-temporal}
       - POSTGRES_PASSWORD=${TEMPORAL_POSTGRES_PASSWORD}
     ports:
       - 25432:5432
     volumes:
       - temporal-db-data:/var/lib/postgresql/data
     healthcheck:
-      test: [ "CMD-SHELL", "pg_isready -d ${TEMPORAL_POSTGRES_DB} -U ${TEMPORAL_POSTGRES_USER}" ]
+      test: [ "CMD-SHELL", "pg_isready -d ${TEMPORAL_POSTGRES_DB:-temporal} -U ${TEMPORAL_POSTGRES_USER:-temporal}" ]
+
       interval: 1s
       timeout: 5s
       retries: 10