Merge pull request #97 from openchatai/refactor/api_flows

using prance to resolve refs, hence this will remove lot of unnecessa…

docker-compose.yml

@@ -78,14 +78,14 @@ services:
     networks:
       - opencopilot_network
 
-  # adminer:
-  #   image: adminer
-  #   ports:
-  #     - "8080:8080"
-  #   depends_on:
-  #     - mysql
-  #   networks:
-  #     - opencopilot_network
+  adminer:
+    image: adminer
+    ports:
+      - "8080:8080"
+    depends_on:
+      - mysql
+    networks:
+      - opencopilot_network
 
   qdrant:
     image: qdrant/qdrant

llm-server/api_caller/planner.py

@@ -245,13 +245,25 @@ def _create_api_controller_agent(
 ) -> AgentExecutor:
     get_llm_chain = LLMChain(llm=llm, prompt=PARSING_GET_PROMPT)
     post_llm_chain = LLMChain(llm=llm, prompt=PARSING_POST_PROMPT)
+    put_llm_chain = LLMChain(llm=llm, prompt=PARSING_PUT_PROMPT)
+    delete_llm_chain = LLMChain(llm=llm, prompt=PARSING_DELETE_PROMPT)
+    path_llm_chain = LLMChain(llm=llm, prompt=PARSING_PATCH_PROMPT)
     tools: List[BaseTool] = [
         RequestsGetToolWithParsing(
             requests_wrapper=requests_wrapper, llm_chain=get_llm_chain
         ),
         RequestsPostToolWithParsing(
             requests_wrapper=requests_wrapper, llm_chain=post_llm_chain
         ),
+        RequestsPutToolWithParsing(
+            requests_wrapper=requests_wrapper, llm_chain=put_llm_chain
+        ),
+        # RequestsDeleteToolWithParsing(
+        #     requests_wrapper=requests_wrapper, llm_chain=delete_llm_chain
+        # ),
+        # RequestsPatchToolWithParsing(
+        #     requests_wrapper=requests_wrapper, llm_chain=path_llm_chain
+        # ),
     ]
     prompt = PromptTemplate(
         template=API_CONTROLLER_PROMPT,
@@ -293,7 +305,7 @@ def _create_and_run_api_controller_agent(plan_str: str) -> str:
             "{method} {route}".format(method=method, route=route.split("?")[0])
             for method, route in matches
         ]
-        endpoint_docs_by_name = {name: docs for name, _, docs in api_spec.endpoints}
+        # endpoint_docs_by_name = {name: docs for name, _, docs in api_spec.endpoints}
         docs_str = ""
         for endpoint_name in endpoint_names:
             found_match = False
@@ -359,5 +371,8 @@ def create_openapi_agent(
         callback_manager=callback_manager,
         verbose=verbose,
         maxIterations=2,
+        early_stopping_method="generate",  # allow one last pass to generate correct response
+        max_execution_time=20,  # kill after 20 seconds
+        handle_parsing_errors=True,
         **(agent_executor_kwargs or {}),
     )

llm-server/app.py

@@ -21,9 +21,9 @@ def handle() -> Response:
     data = request.get_json()
     try:
         response = handle_request(data)
-        return jsonify(response), 200
+        return jsonify(response)
     except Exception as e:
-        return jsonify({"response": str(e)}), 500
+        return jsonify({"response": str(e)})
 
 
 @app.errorhandler(500)