diff --git a/agenthub/README.md b/agenthub/README.md index c2c756a819..d4c80e8214 100644 --- a/agenthub/README.md +++ b/agenthub/README.md @@ -26,6 +26,7 @@ The `state` contains: Here is a list of available Actions, which can be returned by `agent.step()`: - [`CmdRunAction`](../opendevin/action/bash.py) - Runs a command inside a sandboxed terminal - [`CmdKillAction`](../opendevin/action/bash.py) - Kills a background command +- [`IPythonRunCellAction`](../opendevin/action/bash.py) - Execute a block of Python code interactively (in Jupyter notebook) and receives `CmdOutputObservation`. Requires setting up `jupyter` [plugin](../opendevin/sandbox/plugins) as a requirement. - [`FileReadAction`](../opendevin/action/fileop.py) - Reads the content of a file - [`FileWriteAction`](../opendevin/action/fileop.py) - Writes new content to a file - [`BrowseURLAction`](../opendevin/action/browse.py) - Gets the content of a URL @@ -33,6 +34,7 @@ Here is a list of available Actions, which can be returned by `agent.step()`: - [`AddTaskAction`](../opendevin/action/tasks.py) - Adds a subtask to the plan - [`ModifyTaskAction`](../opendevin/action/tasks.py) - Changes the state of a subtask - [`AgentThinkAction`](../opendevin/action/agent.py) - A no-op that allows the agent to add plaintext to the history (as well as the chat log) +- [`AgentTalkAction`](../opendevin/action/agent.py) - A no-op that allows the agent to add plaintext to the history and talk to the user. - [`AgentFinishAction`](../opendevin/action/agent.py) - Stops the control loop, allowing the user to enter a new task You can use `action.to_dict()` and `action_from_dict` to serialize and deserialize actions. diff --git a/agenthub/codeact_agent/codeact_agent.py b/agenthub/codeact_agent/codeact_agent.py index 576e060bfa..59b8ba131c 100644 --- a/agenthub/codeact_agent/codeact_agent.py +++ b/agenthub/codeact_agent/codeact_agent.py @@ -1,54 +1,37 @@ import re from typing import List, Mapping +from agenthub.codeact_agent.prompt import EXAMPLES, SYSTEM_MESSAGE from opendevin.action import ( Action, AgentEchoAction, AgentFinishAction, + AgentTalkAction, CmdRunAction, + IPythonRunCellAction, + NullAction, ) from opendevin.agent import Agent from opendevin.llm.llm import LLM from opendevin.observation import ( AgentMessageObservation, CmdOutputObservation, + IPythonRunCellObservation, + UserMessageObservation, +) +from opendevin.sandbox.plugins import ( + JupyterRequirement, + PluginRequirement, + SWEAgentCommandsRequirement, ) -from opendevin.sandbox.plugins import JupyterRequirement, PluginRequirement from opendevin.state import State -SYSTEM_MESSAGE = """You are a helpful assistant. You will be provided access (as root) to a bash shell to complete user-provided tasks. -You will be able to execute commands in the bash shell, interact with the file system, install packages, and receive the output of your commands. - -DO NOT provide code in ```triple backticks```. Instead, you should execute bash command on behalf of the user by wrapping them with and . -For example: - -You can list the files in the current directory by executing the following command: -ls - -You can also install packages using pip: - pip install numpy - -You can also write a block of code to a file: - -echo "import math -print(math.pi)" > math.py - - -When you are done, execute the following to close the shell and end the conversation: -exit -""" - -INVALID_INPUT_MESSAGE = ( - "I don't understand your input. \n" - 'If you want to execute command, please use YOUR_COMMAND_HERE .\n' - 'If you already completed the task, please exit the shell by generating: exit .' -) - def parse_response(response) -> str: action = response.choices[0].message.content - if '' in action and '' not in action: - action += '' + for lang in ['bash', 'ipython']: + if f'' in action and f'' not in action: + action += f'' return action @@ -58,7 +41,20 @@ class CodeActAgent(Agent): The agent works by passing the model a list of action-observation pairs and prompting the model to take the next step. """ - sandbox_plugins: List[PluginRequirement] = [JupyterRequirement()] + sandbox_plugins: List[PluginRequirement] = [JupyterRequirement(), SWEAgentCommandsRequirement()] + SUPPORTED_ACTIONS = ( + CmdRunAction, + IPythonRunCellAction, + AgentEchoAction, + AgentTalkAction, + NullAction + ) + SUPPORTED_OBSERVATIONS = ( + AgentMessageObservation, + UserMessageObservation, + CmdOutputObservation, + IPythonRunCellObservation + ) def __init__( self, @@ -93,56 +89,76 @@ class CodeActAgent(Agent): assert state.plan.main_goal, 'Expecting instruction to be set' self.messages = [ {'role': 'system', 'content': SYSTEM_MESSAGE}, - {'role': 'user', 'content': state.plan.main_goal}, + { + 'role': 'user', + 'content': ( + f'Here is an example of how you can interact with the environment for task solving:\n{EXAMPLES}\n\n' + f"NOW, LET'S START!\n\n{state.plan.main_goal}" + ) + }, ] updated_info = state.updated_info if updated_info: for prev_action, obs in updated_info: assert isinstance( - prev_action, (CmdRunAction, AgentEchoAction) - ), 'Expecting CmdRunAction or AgentEchoAction for Action' - if isinstance( - obs, AgentMessageObservation - ): # warning message from itself + prev_action, self.SUPPORTED_ACTIONS + ), f'{prev_action.__class__} is not supported (supported: {self.SUPPORTED_ACTIONS})' + # prev_action is already added to self.messages when returned + + # handle observations + assert isinstance( + obs, self.SUPPORTED_OBSERVATIONS + ), f'{obs.__class__} is not supported (supported: {self.SUPPORTED_OBSERVATIONS})' + if isinstance(obs, (AgentMessageObservation, UserMessageObservation)): self.messages.append( {'role': 'user', 'content': obs.content}) + + # User wants to exit + if obs.content.strip() == '/exit': + return AgentFinishAction() elif isinstance(obs, CmdOutputObservation): content = 'OBSERVATION:\n' + obs.content content += f'\n[Command {obs.command_id} finished with exit code {obs.exit_code}]]' self.messages.append({'role': 'user', 'content': content}) + elif isinstance(obs, IPythonRunCellObservation): + content = 'OBSERVATION:\n' + obs.content + self.messages.append({'role': 'user', 'content': content}) else: raise NotImplementedError( f'Unknown observation type: {obs.__class__}' ) + response = self.llm.completion( messages=self.messages, - stop=[''], + stop=[ + '', + '', + ], temperature=0.0 ) action_str: str = parse_response(response) - state.num_of_chars += sum(len(message['content']) - for message in self.messages) + len(action_str) + state.num_of_chars += sum( + len(message['content']) for message in self.messages + ) + len(action_str) self.messages.append({'role': 'assistant', 'content': action_str}) - command = re.search(r'(.*)', action_str, re.DOTALL) - if command is not None: + if bash_command := re.search(r'(.*)', action_str, re.DOTALL): + # remove the command from the action string to get thought + thought = action_str.replace(bash_command.group(0), '').strip() # a command was found - command_group = command.group(1) + command_group = bash_command.group(1).strip() if command_group.strip() == 'exit': return AgentFinishAction() - return CmdRunAction(command=command_group) - # # execute the code - # # TODO: does exit_code get loaded into Message? - # exit_code, observation = self.env.execute(command_group) - # self._history.append(Message(Role.ASSISTANT, observation)) + return CmdRunAction(command=command_group, thought=thought) + elif python_code := re.search(r'(.*)', action_str, re.DOTALL): + # a code block was found + code_group = python_code.group(1).strip() + thought = action_str.replace(python_code.group(0), '').strip() + return IPythonRunCellAction(code=code_group, thought=thought) else: - # we could provide a error message for the model to continue similar to - # https://github.com/xingyaoww/mint-bench/blob/main/mint/envs/general_env.py#L18-L23 - # observation = INVALID_INPUT_MESSAGE - # self._history.append(Message(Role.ASSISTANT, observation)) - return AgentEchoAction( - content=INVALID_INPUT_MESSAGE - ) # warning message to itself + # We assume the LLM is GOOD enough that when it returns pure natural language + # it want to talk to the user + return AgentTalkAction(content=action_str) def search_memory(self, query: str) -> List[str]: raise NotImplementedError('Implement this abstract method') diff --git a/agenthub/codeact_agent/prompt.py b/agenthub/codeact_agent/prompt.py new file mode 100644 index 0000000000..b504a12951 --- /dev/null +++ b/agenthub/codeact_agent/prompt.py @@ -0,0 +1,226 @@ +from opendevin.sandbox.plugins import SWEAgentCommandsRequirement + +_SWEAGENT_BASH_DOCS = '\n'.join( + filter( + lambda x: not x.startswith('submit'), + SWEAgentCommandsRequirement.documentation.split('\n') + ) +) +# _SWEAGENT_BASH_DOCS content below: +""" +open [] - opens the file at the given path in the editor. If line_number is provided, the window will be move to include that line +goto - moves the window to show +scroll_down - moves the window down {WINDOW} lines +scroll_up - moves the window down {WINDOW} lines +create - creates and opens a new file with the given name +search_dir [] - searches for search_term in all files in dir. If dir is not provided, searches in the current directory +search_file [] - searches for search_term in file. If file is not provided, searches in the current open file +find_file [] - finds all files with the given name in dir. If dir is not provided, searches in the current directory +edit : + +end_of_edit - replaces lines through (inclusive) with the given text in the open file. The replacement text is terminated by a line with only end_of_edit on it. All of the will be entered, so make sure your indentation is formatted properly. Python files will be checked for syntax errors after the edit. If the system detects a syntax error, the edit will not be executed. Simply try to edit the file again, but make sure to read the error message and modify the edit command you issue accordingly. Issuing the same command a second time will just lead to the same error message again. +""" + +_COMMAND_DOCS = ( + '\nApart from the standard bash commands, you can also use the following special commands in environment:\n' + f'{_SWEAGENT_BASH_DOCS}' + "Please note that THE EDIT COMMAND REQUIRES PROPER INDENTATION. If you'd like to add the line ' print(x)' you must fully write that out, with all those spaces before the code! Indentation is important and code that is not indented correctly will fail and require fixing before it can be run." +) + +SYSTEM_MESSAGE = f"""A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. +The assistant can interact with an interactive Python (Jupyter Notebook) environment and receive the corresponding output when needed. The code should be enclosed using "" tag, for example: + +print("Hello World!") + +The assistant can execute bash commands on behalf of the user by wrapping them with and . +For example, you can list the files in the current directory by ls . +The assistant should attempt fewer things at a time instead of putting too much commands OR code in one "execute" block. +The assistant can install Python packages through bash by pip install [package needed] and should always import packages and define variables before starting to use them. +The assistant should stop and provide an answer when they have already obtained the answer from the execution result. + +{_COMMAND_DOCS} + +Whenever possible, execute the code for the user using or instead of providing it. +The assistant's response should be concise, but do express their thoughts. +Try to include one of or in each of your responses, unless it is a direct answer to a question OR a message to the user. +""" + +EXAMPLES = """ +--- START OF EXAMPLE --- + +USER: Can you create a list of numbers from 1 to 10, and create a web page to display them at port 5000? + +ASSISTANT: +Sure! Let me write the Python code for starting a web server and save it to a file `app.py`: + +CODE=''' +from flask import Flask +app = Flask(__name__) + +@app.route('/') +def index(): + numbers = list(range(1, 11)) + return str(numbers) + +if __name__ == '__main__': + app.run(port=5000) +''' +with open('app.py', 'w') as f: + f.write(CODE) + + +USER: +Observation: +[Code executed successfully with no output] + +ASSISTANT: +I have created a Python file `app.py` that will display a list of numbers from 1 to 10 when you run it. Let me run the Python file for you: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 121[1]+ Exit 1 python3 app.py > server.log 2>&1 + +ASSISTANT: Looks like the server is running with PID 121 then crashed. Let me check the server log: + +cat server.log + + +USER: +Observation: +Traceback (most recent call last): + File "/workspace/app.py", line 2, in + from flask import Flask +ModuleNotFoundError: No module named 'flask' + +ASSISTANT: +It seems that Flask is not installed. Let me install Flask for you: + +pip install flask + + +USER: +Observation: +Defaulting to user installation because normal site-packages is not writeable +Collecting flask + Using cached flask-3.0.3-py3-none-any.whl (101 kB) +Collecting blinker>=1.6.2 + Using cached blinker-1.7.0-py3-none-any.whl (13 kB) +Collecting Werkzeug>=3.0.0 + Using cached werkzeug-3.0.2-py3-none-any.whl (226 kB) +Collecting click>=8.1.3 + Using cached click-8.1.7-py3-none-any.whl (97 kB) +Collecting itsdangerous>=2.1.2 + Using cached itsdangerous-2.2.0-py3-none-any.whl (16 kB) +Requirement already satisfied: Jinja2>=3.1.2 in /home/opendevin/.local/lib/python3.10/site-packages (from flask) (3.1.3) +Requirement already satisfied: MarkupSafe>=2.0 in /home/opendevin/.local/lib/python3.10/site-packages (from Jinja2>=3.1.2->flask) (2.1.5) +Installing collected packages: Werkzeug, itsdangerous, click, blinker, flask +Successfully installed Werkzeug-3.0.2 blinker-1.7.0 click-8.1.7 flask-3.0.3 itsdangerous-2.2.0 + +ASSISTANT: +Now that Flask is installed, let me run the Python file again: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 124 + +ASSISTANT: +Let me check the server log again: + +cat server.log + + +USER: +Observation: +* Serving Flask app 'app' + * Debug mode: off +WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead. + * Running on http://127.0.0.1:5000 +Press CTRL+C to quit + +ASSISTANT: +The server is running on port 5000 with PID 124. You can access the list of numbers by visiting http://127.0.0.1:5000. If you have any further questions, feel free to ask! + +USER: Now kill the server, make it display the numbers in a table format. + +ASSISTANT: +Sure! Let me stop the server first: + +kill 124 + + +USER: +Observation: +[1]+ Terminated python3 app.py > server.log 2>&1 + +ASSISTANT: +The server has been stopped. Let me open the Python file and modify it. + +open app.py + + +USER: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return str(numbers) +9: +10:if __name__ == '__main__': +11: app.run(port=5000) + +ASSISTANT: +I should edit the file to display the numbers in a table format. Let me do that for you: + + +edit 8:8 <' + ''.join([f'{i}' for i in numbers]) + '' +EOF + + +USER: +Observation: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return '' + ''.join([f'' for i in numbers]) + '
{i}
' +9: +10:if __name__ == '__main__': +11: app.run(port=5000) +File updated. Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary. + +ASSISTANT: +The file has been updated. Let me run the Python file again with the new changes: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 126 + +ASSISTANT: +The server is running on port 5000 with PID 126. You can access the list of numbers in a table format by visiting http://127.0.0.1:5000. Free free to let me know if you have any further requests! + +--- END OF EXAMPLE --- +""" + +INVALID_INPUT_MESSAGE = ( + "I don't understand your input. \n" + 'If you want to execute a bash command, please use YOUR_COMMAND_HERE .\n' + 'If you want to execute a block of Python code, please use YOUR_COMMAND_HERE .\n' +) diff --git a/containers/sandbox/Dockerfile b/containers/sandbox/Dockerfile index 5c0a2c9ea7..1f81645f99 100644 --- a/containers/sandbox/Dockerfile +++ b/containers/sandbox/Dockerfile @@ -27,3 +27,7 @@ RUN mkdir -p -m0755 /var/run/sshd # symlink python3 to python RUN ln -s /usr/bin/python3 /usr/bin/python + +# install basic dependencies for CodeActAgent +RUN pip3 install --upgrade pip +RUN pip3 install jupyterlab notebook jupyter_kernel_gateway flake8 diff --git a/docs/modules/python/opendevin/observation/run.md b/docs/modules/python/opendevin/observation/run.md index a311d40a7f..c0914c70ac 100644 --- a/docs/modules/python/opendevin/observation/run.md +++ b/docs/modules/python/opendevin/observation/run.md @@ -12,3 +12,12 @@ class CmdOutputObservation(Observation) This data class represents the output of a command. +## IPythonRunCellObservation Objects + +```python +@dataclass +class IPythonRunCellObservation(Observation) +``` + +This data class represents the output of a IPythonRunCellAction. + diff --git a/docs/modules/python/opendevin/schema/action.md b/docs/modules/python/opendevin/schema/action.md index 8046dc23e7..0c35c47186 100644 --- a/docs/modules/python/opendevin/schema/action.md +++ b/docs/modules/python/opendevin/schema/action.md @@ -13,9 +13,13 @@ class ActionTypeSchema(BaseModel) Initializes the agent. Only sent by client. +#### USER\_MESSAGE + +Sends a message from the user. Only sent by the client. + #### START -Starts a new development task. Only sent by the client. +Starts a new development task OR send chat from the user. Only sent by the client. #### READ @@ -29,6 +33,10 @@ Writes the content to a file. Runs a command. +#### RUN\_IPYTHON + +Runs a IPython cell. + #### KILL Kills a background command. @@ -45,6 +53,10 @@ Searches long-term memory Allows the agent to make a plan, set a goal, or record thoughts +#### TALK + +Allows the agent to respond to the user. + #### DELEGATE Delegates a task to another agent. diff --git a/docs/modules/python/opendevin/schema/observation.md b/docs/modules/python/opendevin/schema/observation.md index db6a80ab83..c327e61d5f 100644 --- a/docs/modules/python/opendevin/schema/observation.md +++ b/docs/modules/python/opendevin/schema/observation.md @@ -21,6 +21,10 @@ The HTML content of a URL The output of a command +#### RUN\_IPYTHON + +Runs a IPython cell. + #### RECALL The result of a search diff --git a/docs/modules/python/opendevin/schema/task.md b/docs/modules/python/opendevin/schema/task.md index 4f21dc2559..ba1900bcce 100644 --- a/docs/modules/python/opendevin/schema/task.md +++ b/docs/modules/python/opendevin/schema/task.md @@ -17,6 +17,10 @@ Initial state of the task. The task is running. +#### AWAITING\_USER\_INPUT + +The task is awaiting user input. + #### PAUSED The task is paused. diff --git a/frontend/src/components/AgentStatusBar.tsx b/frontend/src/components/AgentStatusBar.tsx index f1ac56aa7b..253164b64e 100644 --- a/frontend/src/components/AgentStatusBar.tsx +++ b/frontend/src/components/AgentStatusBar.tsx @@ -15,6 +15,10 @@ const AgentStatusMap: { [k: string]: { message: string; indicator: string } } = message: "Agent is running task...", indicator: "bg-green-500", }, + [AgentTaskState.AWAITING_USER_INPUT]: { + message: "Agent is awaiting user input...", + indicator: "bg-orange-500", + }, [AgentTaskState.PAUSED]: { message: "Agent has paused.", indicator: "bg-yellow-500", diff --git a/frontend/src/components/ChatInterface.tsx b/frontend/src/components/ChatInterface.tsx index e113244f5c..982ac99cb1 100644 --- a/frontend/src/components/ChatInterface.tsx +++ b/frontend/src/components/ChatInterface.tsx @@ -3,6 +3,7 @@ import { IoMdChatbubbles } from "react-icons/io"; import Markdown from "react-markdown"; import { useSelector } from "react-redux"; import { useTypingEffect } from "#/hooks/useTypingEffect"; +import AgentTaskState from "../types/AgentTaskState"; import { addAssistantMessageToChat, sendChatMessage, @@ -117,6 +118,12 @@ function MessageList(): JSX.Element { function ChatInterface(): JSX.Element { const { initialized } = useSelector((state: RootState) => state.task); + const { curTaskState } = useSelector((state: RootState) => state.agent); + + const onUserMessage = (msg: string) => { + const isNewTask = curTaskState === AgentTaskState.INIT; + sendChatMessage(msg, isNewTask); + }; return (
@@ -125,7 +132,7 @@ function ChatInterface(): JSX.Element { Chat
- + ); } diff --git a/frontend/src/components/modals/load-previous-session/LoadPreviousSessionModal.test.tsx b/frontend/src/components/modals/load-previous-session/LoadPreviousSessionModal.test.tsx index 89d70d39f5..dc6a6516a0 100644 --- a/frontend/src/components/modals/load-previous-session/LoadPreviousSessionModal.test.tsx +++ b/frontend/src/components/modals/load-previous-session/LoadPreviousSessionModal.test.tsx @@ -3,7 +3,7 @@ import { act, render, screen, waitFor } from "@testing-library/react"; import userEvent from "@testing-library/user-event"; import LoadPreviousSessionModal from "./LoadPreviousSessionModal"; import { clearMsgs, fetchMsgs } from "../../../services/session"; -import { sendChatMessageFromEvent } from "../../../services/chatService"; +import { addChatMessageFromEvent } from "../../../services/chatService"; import { handleAssistantMessage } from "../../../services/actions"; import toast from "../../../utils/toast"; @@ -37,7 +37,7 @@ vi.mock("../../../services/session", async (importOriginal) => ({ vi.mock("../../../services/chatService", async (importOriginal) => ({ ...(await importOriginal()), - sendChatMessageFromEvent: vi.fn(), + addChatMessageFromEvent: vi.fn(), })); vi.mock("../../../services/actions", async (importOriginal) => ({ @@ -94,7 +94,7 @@ describe("LoadPreviousSession", () => { await waitFor(() => { expect(fetchMsgs).toHaveBeenCalledTimes(1); - expect(sendChatMessageFromEvent).toHaveBeenCalledTimes(1); + expect(addChatMessageFromEvent).toHaveBeenCalledTimes(1); expect(handleAssistantMessage).toHaveBeenCalledTimes(1); }); // modal should close right after fetching messages @@ -117,7 +117,7 @@ describe("LoadPreviousSession", () => { await waitFor(async () => { await expect(() => fetchMsgs()).rejects.toThrow(); expect(handleAssistantMessage).not.toHaveBeenCalled(); - expect(sendChatMessageFromEvent).not.toHaveBeenCalled(); + expect(addChatMessageFromEvent).not.toHaveBeenCalled(); // error toast should be shown expect(toast.stickyError).toHaveBeenCalledWith( "ws", diff --git a/frontend/src/components/modals/load-previous-session/LoadPreviousSessionModal.tsx b/frontend/src/components/modals/load-previous-session/LoadPreviousSessionModal.tsx index 3d8beef582..2f29f05f10 100644 --- a/frontend/src/components/modals/load-previous-session/LoadPreviousSessionModal.tsx +++ b/frontend/src/components/modals/load-previous-session/LoadPreviousSessionModal.tsx @@ -2,7 +2,7 @@ import React from "react"; import { useTranslation } from "react-i18next"; import { I18nKey } from "#/i18n/declaration"; import { handleAssistantMessage } from "#/services/actions"; -import { sendChatMessageFromEvent } from "#/services/chatService"; +import { addChatMessageFromEvent } from "#/services/chatService"; import { clearMsgs, fetchMsgs } from "#/services/session"; import toast from "#/utils/toast"; import BaseModal from "../base-modal/BaseModal"; @@ -28,7 +28,7 @@ function LoadPreviousSessionModal({ messages.forEach((message) => { if (message.role === "user") { - sendChatMessageFromEvent(message.payload); + addChatMessageFromEvent(message.payload); } if (message.role === "assistant") { diff --git a/frontend/src/components/modals/settings/SettingsForm.tsx b/frontend/src/components/modals/settings/SettingsForm.tsx index 21042fbc22..7387c397d4 100644 --- a/frontend/src/components/modals/settings/SettingsForm.tsx +++ b/frontend/src/components/modals/settings/SettingsForm.tsx @@ -38,7 +38,8 @@ function SettingsForm({ useEffect(() => { if ( curTaskState === AgentTaskState.RUNNING || - curTaskState === AgentTaskState.PAUSED + curTaskState === AgentTaskState.PAUSED || + curTaskState === AgentTaskState.AWAITING_USER_INPUT ) { setDisabled(true); } else { diff --git a/frontend/src/services/actions.ts b/frontend/src/services/actions.ts index f4551a8960..d4484796ca 100644 --- a/frontend/src/services/actions.ts +++ b/frontend/src/services/actions.ts @@ -29,12 +29,24 @@ const messageActions = { [ActionType.THINK]: (message: ActionMessage) => { store.dispatch(appendAssistantMessage(message.args.thought)); }, + [ActionType.TALK]: (message: ActionMessage) => { + store.dispatch(appendAssistantMessage(message.args.content)); + }, [ActionType.FINISH]: (message: ActionMessage) => { store.dispatch(appendAssistantMessage(message.message)); }, [ActionType.RUN]: (message: ActionMessage) => { + if (message.args.thought) { + store.dispatch(appendAssistantMessage(message.args.thought)); + } store.dispatch(appendInput(message.args.command)); }, + [ActionType.RUN_IPYTHON]: (message: ActionMessage) => { + if (message.args.thought) { + store.dispatch(appendAssistantMessage(message.args.thought)); + } + store.dispatch(appendInput(message.args.code)); + }, [ActionType.ADD_TASK]: () => { getPlan().then((fetchedPlan) => store.dispatch(setPlan(fetchedPlan))); }, diff --git a/frontend/src/services/chatService.ts b/frontend/src/services/chatService.ts index 18d4042f2a..8c01b0079f 100644 --- a/frontend/src/services/chatService.ts +++ b/frontend/src/services/chatService.ts @@ -11,14 +11,19 @@ import { SocketMessage } from "#/types/ResponseType"; import { ActionMessage } from "#/types/Message"; import Socket from "./socket"; -export function sendChatMessage(message: string): void { +export function sendChatMessage(message: string, isTask: boolean = true): void { store.dispatch(appendUserMessage(message)); - const event = { action: ActionType.START, args: { task: message } }; + let event; + if (isTask) { + event = { action: ActionType.START, args: { task: message } }; + } else { + event = { action: ActionType.USER_MESSAGE, args: { message } }; + } const eventString = JSON.stringify(event); Socket.send(eventString); } -export function sendChatMessageFromEvent(event: string | SocketMessage): void { +export function addChatMessageFromEvent(event: string | SocketMessage): void { try { let data: ActionMessage; if (typeof event === "string") { diff --git a/frontend/src/services/observations.ts b/frontend/src/services/observations.ts index 39cafafd93..c5a5cbab7b 100644 --- a/frontend/src/services/observations.ts +++ b/frontend/src/services/observations.ts @@ -10,6 +10,10 @@ export function handleObservationMessage(message: ObservationMessage) { case ObservationType.RUN: store.dispatch(appendOutput(message.content)); break; + case ObservationType.RUN_IPYTHON: + // FIXME: render this as markdown + store.dispatch(appendOutput(message.content)); + break; case ObservationType.BROWSE: if (message.extras?.screenshot) { store.dispatch(setScreenshotSrc(message.extras.screenshot)); diff --git a/frontend/src/types/ActionType.tsx b/frontend/src/types/ActionType.tsx index 1d88324eb0..1862a13f2d 100644 --- a/frontend/src/types/ActionType.tsx +++ b/frontend/src/types/ActionType.tsx @@ -2,7 +2,10 @@ enum ActionType { // Initializes the agent. Only sent by client. INIT = "initialize", - // Starts a new development task. Only sent by the client. + // Sends a message from the user + USER_MESSAGE = "user_message", + + // Starts a new development task START = "start", // Reads the contents of a file. @@ -14,6 +17,9 @@ enum ActionType { // Runs a command. RUN = "run", + // Runs a IPython command. + RUN_IPYTHON = "run_ipython", + // Kills a background command. KILL = "kill", @@ -26,6 +32,9 @@ enum ActionType { // Allows the agent to make a plan, set a goal, or record thoughts. THINK = "think", + // Allows the agent to respond to the user. Only sent by the agent. + TALK = "talk", + // If you're absolutely certain that you've completed your task and have tested your work, // use the finish action to stop working. FINISH = "finish", diff --git a/frontend/src/types/AgentTaskState.tsx b/frontend/src/types/AgentTaskState.tsx index d7aa9c80bd..b5cd1d5a52 100644 --- a/frontend/src/types/AgentTaskState.tsx +++ b/frontend/src/types/AgentTaskState.tsx @@ -1,6 +1,7 @@ enum AgentTaskState { INIT = "init", RUNNING = "running", + AWAITING_USER_INPUT = "awaiting_user_input", PAUSED = "paused", STOPPED = "stopped", FINISHED = "finished", diff --git a/frontend/src/types/ObservationType.tsx b/frontend/src/types/ObservationType.tsx index b86bd58919..6ba8bda578 100644 --- a/frontend/src/types/ObservationType.tsx +++ b/frontend/src/types/ObservationType.tsx @@ -8,6 +8,9 @@ enum ObservationType { // The output of a command RUN = "run", + // The output of an IPython command + RUN_IPYTHON = "run_ipython", + // The result of a search RECALL = "recall", diff --git a/opendevin/action/__init__.py b/opendevin/action/__init__.py index be79bbfe82..9e3d939d3f 100644 --- a/opendevin/action/__init__.py +++ b/opendevin/action/__init__.py @@ -5,10 +5,11 @@ from .agent import ( AgentFinishAction, AgentRecallAction, AgentSummarizeAction, + AgentTalkAction, AgentThinkAction, ) from .base import Action, NullAction -from .bash import CmdKillAction, CmdRunAction +from .bash import CmdKillAction, CmdRunAction, IPythonRunCellAction from .browse import BrowseURLAction from .fileop import FileReadAction, FileWriteAction from .github import GitHubPushAction @@ -17,11 +18,13 @@ from .tasks import AddTaskAction, ModifyTaskAction actions = ( CmdKillAction, CmdRunAction, + IPythonRunCellAction, BrowseURLAction, FileReadAction, FileWriteAction, AgentRecallAction, AgentThinkAction, + AgentTalkAction, AgentFinishAction, AgentDelegateAction, AddTaskAction, @@ -61,10 +64,12 @@ __all__ = [ 'FileWriteAction', 'AgentRecallAction', 'AgentThinkAction', + 'AgentTalkAction', 'AgentFinishAction', 'AgentDelegateAction', 'AgentEchoAction', 'AgentSummarizeAction', 'AddTaskAction', 'ModifyTaskAction', + 'IPythonRunCellAction' ] diff --git a/opendevin/action/agent.py b/opendevin/action/agent.py index 4efff3db8d..063d095bd7 100644 --- a/opendevin/action/agent.py +++ b/opendevin/action/agent.py @@ -18,6 +18,7 @@ if TYPE_CHECKING: @dataclass class AgentRecallAction(ExecutableAction): query: str + thought: str = '' action: str = ActionType.RECALL async def run(self, controller: 'AgentController') -> AgentRecallObservation: @@ -44,6 +45,22 @@ class AgentThinkAction(NotExecutableAction): return self.thought +@dataclass +class AgentTalkAction(NotExecutableAction): + content: str + action: str = ActionType.TALK + + async def run(self, controller: 'AgentController') -> 'Observation': + raise NotImplementedError + + @property + def message(self) -> str: + return self.content + + def __str__(self) -> str: + return self.content + + @dataclass class AgentEchoAction(ExecutableAction): content: str @@ -70,6 +87,7 @@ class AgentSummarizeAction(NotExecutableAction): @dataclass class AgentFinishAction(NotExecutableAction): outputs: Dict = field(default_factory=dict) + thought: str = '' action: str = ActionType.FINISH async def run(self, controller: 'AgentController') -> 'Observation': @@ -84,6 +102,7 @@ class AgentFinishAction(NotExecutableAction): class AgentDelegateAction(ExecutableAction): agent: str inputs: dict + thought: str = '' action: str = ActionType.DELEGATE async def run(self, controller: 'AgentController') -> 'Observation': diff --git a/opendevin/action/bash.py b/opendevin/action/bash.py index 7dc2c4ff27..160a795491 100644 --- a/opendevin/action/bash.py +++ b/opendevin/action/bash.py @@ -1,7 +1,10 @@ +import os +import pathlib from dataclasses import dataclass from typing import TYPE_CHECKING -from opendevin.schema import ActionType +from opendevin import config +from opendevin.schema import ActionType, ConfigType from .base import ExecutableAction @@ -9,11 +12,14 @@ if TYPE_CHECKING: from opendevin.controller import AgentController from opendevin.observation import CmdOutputObservation, Observation +from opendevin.observation import IPythonRunCellObservation + @dataclass class CmdRunAction(ExecutableAction): command: str background: bool = False + thought: str = '' action: str = ActionType.RUN async def run(self, controller: 'AgentController') -> 'Observation': @@ -23,10 +29,18 @@ class CmdRunAction(ExecutableAction): def message(self) -> str: return f'Running command: {self.command}' + def __str__(self) -> str: + ret = '**CmdRunAction**\n' + if self.thought: + ret += f'THOUGHT:{self.thought}\n' + ret += f'COMMAND:\n{self.command}' + return ret + @dataclass class CmdKillAction(ExecutableAction): id: int + thought: str = '' action: str = ActionType.KILL async def run(self, controller: 'AgentController') -> 'CmdOutputObservation': @@ -35,3 +49,48 @@ class CmdKillAction(ExecutableAction): @property def message(self) -> str: return f'Killing command: {self.id}' + + def __str__(self) -> str: + return f'**CmdKillAction**\n{self.id}' + + +@dataclass +class IPythonRunCellAction(ExecutableAction): + code: str + thought: str = '' + action: str = ActionType.RUN_IPYTHON + + async def run(self, controller: 'AgentController') -> 'IPythonRunCellObservation': + # echo "import math" | execute_cli + # write code to a temporary file and pass it to `execute_cli` via stdin + tmp_filepath = os.path.join( + config.get(ConfigType.WORKSPACE_BASE), + '.tmp', '.ipython_execution_tmp.py' + ) + pathlib.Path(os.path.dirname(tmp_filepath)).mkdir(parents=True, exist_ok=True) + with open(tmp_filepath, 'w') as tmp_file: + tmp_file.write(self.code) + + tmp_filepath_inside_sandbox = os.path.join( + config.get(ConfigType.WORKSPACE_MOUNT_PATH_IN_SANDBOX), + '.tmp', '.ipython_execution_tmp.py' + ) + obs = controller.action_manager.run_command( + f'execute_cli < {tmp_filepath_inside_sandbox}', + background=False + ) + return IPythonRunCellObservation( + content=obs.content, + code=self.code + ) + + def __str__(self) -> str: + ret = '**IPythonRunCellAction**\n' + if self.thought: + ret += f'THOUGHT:{self.thought}\n' + ret += f'CODE:\n{self.code}' + return ret + + @property + def message(self) -> str: + return f'Running Python code interactively: {self.code}' diff --git a/opendevin/action/browse.py b/opendevin/action/browse.py index 3c24e04bf2..8ed1525289 100644 --- a/opendevin/action/browse.py +++ b/opendevin/action/browse.py @@ -17,6 +17,7 @@ if TYPE_CHECKING: @dataclass class BrowseURLAction(ExecutableAction): url: str + thought: str = '' action: str = ActionType.BROWSE async def run(self, controller: 'AgentController') -> BrowserOutputObservation: # type: ignore diff --git a/opendevin/action/fileop.py b/opendevin/action/fileop.py index 343a154e9f..4e630c9383 100644 --- a/opendevin/action/fileop.py +++ b/opendevin/action/fileop.py @@ -50,7 +50,7 @@ class FileReadAction(ExecutableAction): path: str start: int = 0 end: int = -1 - thoughts: str = '' + thought: str = '' action: str = ActionType.READ def _read_lines(self, all_lines: list[str]): @@ -100,7 +100,7 @@ class FileWriteAction(ExecutableAction): content: str start: int = 0 end: int = -1 - thoughts: str = '' + thought: str = '' action: str = ActionType.WRITE def _insert_lines(self, to_insert: list[str], original: list[str]): diff --git a/opendevin/action/tasks.py b/opendevin/action/tasks.py index f102340300..5eff530e66 100644 --- a/opendevin/action/tasks.py +++ b/opendevin/action/tasks.py @@ -15,6 +15,7 @@ class AddTaskAction(ExecutableAction): parent: str goal: str subtasks: list = field(default_factory=list) + thought: str = '' action: str = ActionType.ADD_TASK async def run(self, controller: 'AgentController') -> NullObservation: # type: ignore @@ -31,6 +32,7 @@ class AddTaskAction(ExecutableAction): class ModifyTaskAction(ExecutableAction): id: str state: str + thought: str = '' action: str = ActionType.MODIFY_TASK async def run(self, controller: 'AgentController') -> NullObservation: # type: ignore @@ -47,6 +49,7 @@ class ModifyTaskAction(ExecutableAction): class TaskStateChangedAction(NotExecutableAction): """Fake action, just to notify the client that a task state has changed.""" task_state: str + thought: str = '' action: str = ActionType.CHANGE_TASK_STATE @property diff --git a/opendevin/config.py b/opendevin/config.py index 5b38f4314a..1cb3733d90 100644 --- a/opendevin/config.py +++ b/opendevin/config.py @@ -171,7 +171,6 @@ def finalize_config(): if config.get(ConfigType.WORKSPACE_MOUNT_PATH) is None: config[ConfigType.WORKSPACE_MOUNT_PATH] = config.get(ConfigType.WORKSPACE_BASE) - finalize_config() diff --git a/opendevin/controller/agent_controller.py b/opendevin/controller/agent_controller.py index dff0eb0d50..c4c23c11b7 100644 --- a/opendevin/controller/agent_controller.py +++ b/opendevin/controller/agent_controller.py @@ -1,11 +1,13 @@ import asyncio from typing import Callable, List, Type +from agenthub.codeact_agent.codeact_agent import CodeActAgent from opendevin import config from opendevin.action import ( Action, AgentDelegateAction, AgentFinishAction, + AgentTalkAction, NullAction, ) from opendevin.action.tasks import TaskStateChangedAction @@ -23,8 +25,10 @@ from opendevin.observation import ( AgentErrorObservation, NullObservation, Observation, + UserMessageObservation, ) from opendevin.plan import Plan +from opendevin.sandbox import DockerSSHBox from opendevin.schema import TaskState from opendevin.schema.config import ConfigType from opendevin.state import State @@ -64,6 +68,11 @@ class AgentController: # Initialize agent-required plugins for sandbox (if any) self.action_manager.init_sandbox_plugins(agent.sandbox_plugins) + if isinstance(agent, CodeActAgent) and not isinstance(self.action_manager.sandbox, DockerSSHBox): + logger.warning('CodeActAgent requires DockerSSHBox as sandbox! Using other sandbox that are not stateful (LocalBox, DockerExecBox) will not work properly.') + + self._await_user_message_queue: asyncio.Queue = asyncio.Queue() + def update_state_for_step(self, i): if self.state is None: return @@ -174,6 +183,36 @@ class AgentController: async def notify_task_state_changed(self): await self._run_callbacks(TaskStateChangedAction(self._task_state)) + async def add_user_message(self, message: UserMessageObservation): + if self.state is None: + return + + if self._task_state == TaskState.AWAITING_USER_INPUT: + self._await_user_message_queue.put_nowait(message) + + # set the task state to running + self._task_state = TaskState.RUNNING + await self.notify_task_state_changed() + + elif self._task_state == TaskState.RUNNING: + self.add_history(NullAction(), message) + + else: + raise ValueError(f'Task (state: {self._task_state}) is not in a state to add user message') + + async def wait_for_user_input(self) -> UserMessageObservation: + self._task_state = TaskState.AWAITING_USER_INPUT + await self.notify_task_state_changed() + # wait for the next user message + if len(self.callbacks) == 0: + logger.info('Use STDIN to request user message as no callbacks are registered', extra={'msg_type': 'INFO'}) + message = input('Request user input [type /exit to stop interaction] >> ') + user_message_observation = UserMessageObservation(message) + else: + user_message_observation = await self._await_user_message_queue.get() + self._await_user_message_queue.task_done() + return user_message_observation + async def start_delegate(self, action: AgentDelegateAction): AgentCls: Type[Agent] = Agent.get_cls(action.agent) agent = AgentCls(llm=self.agent.llm) @@ -201,7 +240,8 @@ class AgentController: return False logger.info(f'STEP {i}', extra={'msg_type': 'STEP'}) - logger.info(self.state.plan.main_goal, extra={'msg_type': 'PLAN'}) + if i == 0: + logger.info(self.state.plan.main_goal, extra={'msg_type': 'PLAN'}) if self.state.num_of_chars > self.max_chars: raise MaxCharsExceedError(self.state.num_of_chars, self.max_chars) @@ -226,6 +266,14 @@ class AgentController: await self._run_callbacks(action) + # whether to await for user messages + if isinstance(action, AgentTalkAction): + # await for the next user messages + user_message_observation = await self.wait_for_user_input() + logger.info(user_message_observation, extra={'msg_type': 'OBSERVATION'}) + self.add_history(action, user_message_observation) + return False + finished = isinstance(action, AgentFinishAction) if finished: self.state.outputs = action.outputs # type: ignore[attr-defined] diff --git a/opendevin/observation/__init__.py b/opendevin/observation/__init__.py index 2a4dc83dca..174e90bc9c 100644 --- a/opendevin/observation/__init__.py +++ b/opendevin/observation/__init__.py @@ -5,7 +5,7 @@ from .error import AgentErrorObservation from .files import FileReadObservation, FileWriteObservation from .message import AgentMessageObservation, UserMessageObservation from .recall import AgentRecallObservation -from .run import CmdOutputObservation +from .run import CmdOutputObservation, IPythonRunCellObservation observations = ( CmdOutputObservation, @@ -40,6 +40,7 @@ __all__ = [ 'Observation', 'NullObservation', 'CmdOutputObservation', + 'IPythonRunCellObservation', 'BrowserOutputObservation', 'FileReadObservation', 'FileWriteObservation', diff --git a/opendevin/observation/run.py b/opendevin/observation/run.py index aaf5476e87..8de5fbac5c 100644 --- a/opendevin/observation/run.py +++ b/opendevin/observation/run.py @@ -23,3 +23,21 @@ class CmdOutputObservation(Observation): @property def message(self) -> str: return f'Command `{self.command}` executed with exit code {self.exit_code}.' + + +@dataclass +class IPythonRunCellObservation(Observation): + """ + This data class represents the output of a IPythonRunCellAction. + """ + + code: str + observation: str = ObservationType.RUN_IPYTHON + + @property + def error(self) -> bool: + return False # IPython cells do not return exit codes + + @property + def message(self) -> str: + return 'Coded executed in IPython cell.' diff --git a/opendevin/sandbox/plugins/jupyter/execute_server b/opendevin/sandbox/plugins/jupyter/execute_server index 3362b9cc02..828ac520e8 100755 --- a/opendevin/sandbox/plugins/jupyter/execute_server +++ b/opendevin/sandbox/plugins/jupyter/execute_server @@ -188,7 +188,7 @@ class JupyterKernel: if 'image/png' in msg['content']['data']: # use markdone to display image (in case of large image) # outputs.append(f"\n\n") - outputs.append(f"![image](data:image/png;base64,{msg['content']['data']['image / png']})") + outputs.append(f"![image](data:image/png;base64,{msg['content']['data']['image/png']})") elif msg_type == 'execute_reply': execution_done = True diff --git a/opendevin/schema/action.py b/opendevin/schema/action.py index e8962fb6ce..51d21d4c83 100644 --- a/opendevin/schema/action.py +++ b/opendevin/schema/action.py @@ -10,8 +10,12 @@ class ActionTypeSchema(BaseModel): """Initializes the agent. Only sent by client. """ + USER_MESSAGE: str = Field(default='user_message') + """Sends a message from the user. Only sent by the client. + """ + START: str = Field(default='start') - """Starts a new development task. Only sent by the client. + """Starts a new development task OR send chat from the user. Only sent by the client. """ READ: str = Field(default='read') @@ -26,6 +30,10 @@ class ActionTypeSchema(BaseModel): """Runs a command. """ + RUN_IPYTHON: str = Field(default='run_ipython') + """Runs a IPython cell. + """ + KILL: str = Field(default='kill') """Kills a background command. """ @@ -42,6 +50,10 @@ class ActionTypeSchema(BaseModel): """Allows the agent to make a plan, set a goal, or record thoughts """ + TALK: str = Field(default='talk') + """Allows the agent to respond to the user. + """ + DELEGATE: str = Field(default='delegate') """Delegates a task to another agent. """ diff --git a/opendevin/schema/observation.py b/opendevin/schema/observation.py index 6d4a1290fa..e800e3eb86 100644 --- a/opendevin/schema/observation.py +++ b/opendevin/schema/observation.py @@ -20,6 +20,10 @@ class ObservationTypeSchema(BaseModel): """The output of a command """ + RUN_IPYTHON: str = Field(default='run_ipython') + """Runs a IPython cell. + """ + RECALL: str = Field(default='recall') """The result of a search """ diff --git a/opendevin/schema/task.py b/opendevin/schema/task.py index 010af9b188..cde2eb7d04 100644 --- a/opendevin/schema/task.py +++ b/opendevin/schema/task.py @@ -10,6 +10,10 @@ class TaskState(str, Enum): """The task is running. """ + AWAITING_USER_INPUT = 'awaiting_user_input' + """The task is awaiting user input. + """ + PAUSED = 'paused' """The task is paused. """ diff --git a/opendevin/server/agent/agent.py b/opendevin/server/agent/agent.py index 2290fda407..ae384e5f70 100644 --- a/opendevin/server/agent/agent.py +++ b/opendevin/server/agent/agent.py @@ -94,6 +94,8 @@ class AgentUnit: await self.create_controller(data) case ActionType.START: await self.start_task(data) + case ActionType.USER_MESSAGE: + await self.send_user_message(data) case ActionType.CHANGE_TASK_STATE: task_state_action = data.get('args', {}).get('task_state_action', None) if task_state_action is None: @@ -177,23 +179,27 @@ class AgentUnit: Args: start_event: The start event data. """ - if 'task' not in start_event['args']: - await self.send_error('No task specified') - return - await self.send_message('Starting new task...') task = start_event['args']['task'] if self.controller is None: await self.send_error('No agent started. Please wait a second...') return try: - if self.agent_task: - self.agent_task.cancel() + assert not self.agent_task, 'Agent task already running' self.agent_task = asyncio.create_task( self.controller.start(task), name='agent start task loop' ) except Exception as e: await self.send_error(f'Error during task loop: {e}') + async def send_user_message(self, data: dict): + if not self.agent_task or not self.controller: + await self.send_error('No agent started.') + return + + await self.controller.add_user_message( + UserMessageObservation(data['args']['message']) + ) + async def set_task_state(self, new_state_action: TaskStateAction): """Sets the state of the agent task.""" if self.controller is None: diff --git a/tests/integration/README.md b/tests/integration/README.md index 3c13821fe0..21befd6925 100644 --- a/tests/integration/README.md +++ b/tests/integration/README.md @@ -89,12 +89,15 @@ mkdir workspace poetry run python ./opendevin/main.py -i 10 -t "Write a shell script 'hello.sh' that prints 'hello'." -c "MonologueAgent" -d "./workspace" ``` +**NOTE**: If your agent decide to support user-agent interaction via natural language (e.g., you will prompted to enter user resposes when running the above `main.py` command), you should create a file named `tests/integration/mock///user_responses.log` containing all the responses in order you provided to the agent, delimited by newline ('\n'). This will be used to mock the STDIN during testing. + After running the above commands, you should be able to locate the real prompts and responses logged. The log folder follows `logs/llm/%y-%m-%d_%H-%M.log` format. Now, move all files under that folder to `tests/integration/mock//` folder. For example, moving all files from `logs/llm/24-04-23_21-55/` folder to `tests/integration/mock/MonologueAgent/test_write_simple_script` folder. + That's it, you are good to go! When you launch an integration test, mock responses are loaded and used to replace a real LLM, so that we get deterministic and consistent behavior, and most importantly, without spending real diff --git a/tests/integration/conftest.py b/tests/integration/conftest.py index 92042f27fb..4e4946958d 100644 --- a/tests/integration/conftest.py +++ b/tests/integration/conftest.py @@ -1,3 +1,4 @@ +import io import os import re from functools import partial @@ -53,6 +54,28 @@ def get_mock_response(test_name, messages): return resp_file.read() +def mock_user_response(*args, test_name, **kwargs): + """The agent will ask for user input using `input()` when calling `asyncio.run(main(task))`. + This function mocks the user input by providing the response from the mock response file. + + It will read the `user_responses.log` file in the test directory and set as + STDIN input for the agent to read. + """ + user_response_file = os.path.join( + script_dir, + 'mock', + os.environ.get('AGENT'), + test_name, + 'user_responses.log' + ) + if not os.path.exists(user_response_file): + return '' + with open(user_response_file, 'r') as f: + ret = f.read().rstrip() + ret += '\n' + return ret + + def mock_completion(*args, test_name, **kwargs): messages = kwargs['messages'] message_str = '' @@ -67,4 +90,11 @@ def mock_completion(*args, test_name, **kwargs): @pytest.fixture(autouse=True) def patch_completion(monkeypatch, request): test_name = request.node.name + # Mock LLM completion monkeypatch.setattr('opendevin.llm.llm.litellm_completion', partial(mock_completion, test_name=test_name)) + + # Mock user input (only for tests that have user_responses.log) + user_responses_str = mock_user_response(test_name=test_name) + if user_responses_str: + user_responses = io.StringIO(user_responses_str) + monkeypatch.setattr('sys.stdin', user_responses) diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_001.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_001.log new file mode 100644 index 0000000000..1287212b6f --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_001.log @@ -0,0 +1,215 @@ + + +---------- + +A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. +The assistant can interact with an interactive Python (Jupyter Notebook) environment and receive the corresponding output when needed. The code should be enclosed using "" tag, for example: + +print("Hello World!") + +The assistant can execute bash commands on behalf of the user by wrapping them with and . +For example, you can list the files in the current directory by ls . +The assistant should attempt fewer things at a time instead of putting too much commands OR code in one "execute" block. +The assistant can install Python packages through bash by pip install [package needed] and should always import packages and define variables before starting to use them. +The assistant should stop and provide an answer when they have already obtained the answer from the execution result. + + +Apart from the standard bash commands, you can also use the following special commands in environment: +open [] - opens the file at the given path in the editor. If line_number is provided, the window will be move to include that line +goto - moves the window to show +scroll_down - moves the window down {WINDOW} lines +scroll_up - moves the window down {WINDOW} lines +create - creates and opens a new file with the given name +search_dir [] - searches for search_term in all files in dir. If dir is not provided, searches in the current directory +search_file [] - searches for search_term in file. If file is not provided, searches in the current open file +find_file [] - finds all files with the given name in dir. If dir is not provided, searches in the current directory +edit : + +end_of_edit - replaces lines through (inclusive) with the given text in the open file. The replacement text is terminated by a line with only end_of_edit on it. All of the will be entered, so make sure your indentation is formatted properly. Python files will be checked for syntax errors after the edit. If the system detects a syntax error, the edit will not be executed. Simply try to edit the file again, but make sure to read the error message and modify the edit command you issue accordingly. Issuing the same command a second time will just lead to the same error message again. +Please note that THE EDIT COMMAND REQUIRES PROPER INDENTATION. If you'd like to add the line ' print(x)' you must fully write that out, with all those spaces before the code! Indentation is important and code that is not indented correctly will fail and require fixing before it can be run. + +Whenever possible, execute the code for the user using or instead of providing it. +The assistant's response should be concise, but do express their thoughts. +Try to include one of or in each of your responses, unless it is a direct answer to a question OR a message to the user. + + +---------- + +Here is an example of how you can interact with the environment for task solving: + +--- START OF EXAMPLE --- + +USER: Can you create a list of numbers from 1 to 10, and create a web page to display them at port 5000? + +ASSISTANT: +Sure! Let me write the Python code for starting a web server and save it to a file `app.py`: + +CODE=''' +from flask import Flask +app = Flask(__name__) + +@app.route('/') +def index(): + numbers = list(range(1, 11)) + return str(numbers) + +if __name__ == '__main__': + app.run(port=5000) +''' +with open('app.py', 'w') as f: + f.write(CODE) + + +USER: +Observation: +[Code executed successfully with no output] + +ASSISTANT: +I have created a Python file `app.py` that will display a list of numbers from 1 to 10 when you run it. Let me run the Python file for you: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 121[1]+ Exit 1 python3 app.py > server.log 2>&1 + +ASSISTANT: Looks like the server is running with PID 121 then crashed. Let me check the server log: + +cat server.log + + +USER: +Observation: +Traceback (most recent call last): + File "/workspace/app.py", line 2, in + from flask import Flask +ModuleNotFoundError: No module named 'flask' + +ASSISTANT: +It seems that Flask is not installed. Let me install Flask for you: + +pip install flask + + +USER: +Observation: +Defaulting to user installation because normal site-packages is not writeable +Collecting flask + Using cached flask-3.0.3-py3-none-any.whl (101 kB) +Collecting blinker>=1.6.2 + Using cached blinker-1.7.0-py3-none-any.whl (13 kB) +Collecting Werkzeug>=3.0.0 + Using cached werkzeug-3.0.2-py3-none-any.whl (226 kB) +Collecting click>=8.1.3 + Using cached click-8.1.7-py3-none-any.whl (97 kB) +Collecting itsdangerous>=2.1.2 + Using cached itsdangerous-2.2.0-py3-none-any.whl (16 kB) +Requirement already satisfied: Jinja2>=3.1.2 in /home/opendevin/.local/lib/python3.10/site-packages (from flask) (3.1.3) +Requirement already satisfied: MarkupSafe>=2.0 in /home/opendevin/.local/lib/python3.10/site-packages (from Jinja2>=3.1.2->flask) (2.1.5) +Installing collected packages: Werkzeug, itsdangerous, click, blinker, flask +Successfully installed Werkzeug-3.0.2 blinker-1.7.0 click-8.1.7 flask-3.0.3 itsdangerous-2.2.0 + +ASSISTANT: +Now that Flask is installed, let me run the Python file again: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 124 + +ASSISTANT: +Let me check the server log again: + +cat server.log + + +USER: +Observation: +* Serving Flask app 'app' + * Debug mode: off +WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead. + * Running on http://127.0.0.1:5000 +Press CTRL+C to quit + +ASSISTANT: +The server is running on port 5000 with PID 124. You can access the list of numbers by visiting http://127.0.0.1:5000. If you have any further questions, feel free to ask! + +USER: Now kill the server, make it display the numbers in a table format. + +ASSISTANT: +Sure! Let me stop the server first: + +kill 124 + + +USER: +Observation: +[1]+ Terminated python3 app.py > server.log 2>&1 + +ASSISTANT: +The server has been stopped. Let me open the Python file and modify it. + +open app.py + + +USER: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return str(numbers) +9: +10:if __name__ == '__main__': +11: app.run(port=5000) + +ASSISTANT: +I should edit the file to display the numbers in a table format. Let me do that for you: + + +edit 8:8 <' + ''.join([f'{i}' for i in numbers]) + '' +EOF + + +USER: +Observation: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return '' + ''.join([f'' for i in numbers]) + '
{i}
' +9: +10:if __name__ == '__main__': +11: app.run(port=5000) +File updated. Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary. + +ASSISTANT: +The file has been updated. Let me run the Python file again with the new changes: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 126 + +ASSISTANT: +The server is running on port 5000 with PID 126. You can access the list of numbers in a table format by visiting http://127.0.0.1:5000. Free free to let me know if you have any further requests! + +--- END OF EXAMPLE --- + + +NOW, LET'S START! + +Write a shell script 'hello.sh' that prints 'hello'. diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_002.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_002.log new file mode 100644 index 0000000000..f4ad1e93d9 --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_002.log @@ -0,0 +1,229 @@ + + +---------- + +A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. +The assistant can interact with an interactive Python (Jupyter Notebook) environment and receive the corresponding output when needed. The code should be enclosed using "" tag, for example: + +print("Hello World!") + +The assistant can execute bash commands on behalf of the user by wrapping them with and . +For example, you can list the files in the current directory by ls . +The assistant should attempt fewer things at a time instead of putting too much commands OR code in one "execute" block. +The assistant can install Python packages through bash by pip install [package needed] and should always import packages and define variables before starting to use them. +The assistant should stop and provide an answer when they have already obtained the answer from the execution result. + + +Apart from the standard bash commands, you can also use the following special commands in environment: +open [] - opens the file at the given path in the editor. If line_number is provided, the window will be move to include that line +goto - moves the window to show +scroll_down - moves the window down {WINDOW} lines +scroll_up - moves the window down {WINDOW} lines +create - creates and opens a new file with the given name +search_dir [] - searches for search_term in all files in dir. If dir is not provided, searches in the current directory +search_file [] - searches for search_term in file. If file is not provided, searches in the current open file +find_file [] - finds all files with the given name in dir. If dir is not provided, searches in the current directory +edit : + +end_of_edit - replaces lines through (inclusive) with the given text in the open file. The replacement text is terminated by a line with only end_of_edit on it. All of the will be entered, so make sure your indentation is formatted properly. Python files will be checked for syntax errors after the edit. If the system detects a syntax error, the edit will not be executed. Simply try to edit the file again, but make sure to read the error message and modify the edit command you issue accordingly. Issuing the same command a second time will just lead to the same error message again. +Please note that THE EDIT COMMAND REQUIRES PROPER INDENTATION. If you'd like to add the line ' print(x)' you must fully write that out, with all those spaces before the code! Indentation is important and code that is not indented correctly will fail and require fixing before it can be run. + +Whenever possible, execute the code for the user using or instead of providing it. +The assistant's response should be concise, but do express their thoughts. +Try to include one of or in each of your responses, unless it is a direct answer to a question OR a message to the user. + + +---------- + +Here is an example of how you can interact with the environment for task solving: + +--- START OF EXAMPLE --- + +USER: Can you create a list of numbers from 1 to 10, and create a web page to display them at port 5000? + +ASSISTANT: +Sure! Let me write the Python code for starting a web server and save it to a file `app.py`: + +CODE=''' +from flask import Flask +app = Flask(__name__) + +@app.route('/') +def index(): + numbers = list(range(1, 11)) + return str(numbers) + +if __name__ == '__main__': + app.run(port=5000) +''' +with open('app.py', 'w') as f: + f.write(CODE) + + +USER: +Observation: +[Code executed successfully with no output] + +ASSISTANT: +I have created a Python file `app.py` that will display a list of numbers from 1 to 10 when you run it. Let me run the Python file for you: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 121[1]+ Exit 1 python3 app.py > server.log 2>&1 + +ASSISTANT: Looks like the server is running with PID 121 then crashed. Let me check the server log: + +cat server.log + + +USER: +Observation: +Traceback (most recent call last): + File "/workspace/app.py", line 2, in + from flask import Flask +ModuleNotFoundError: No module named 'flask' + +ASSISTANT: +It seems that Flask is not installed. Let me install Flask for you: + +pip install flask + + +USER: +Observation: +Defaulting to user installation because normal site-packages is not writeable +Collecting flask + Using cached flask-3.0.3-py3-none-any.whl (101 kB) +Collecting blinker>=1.6.2 + Using cached blinker-1.7.0-py3-none-any.whl (13 kB) +Collecting Werkzeug>=3.0.0 + Using cached werkzeug-3.0.2-py3-none-any.whl (226 kB) +Collecting click>=8.1.3 + Using cached click-8.1.7-py3-none-any.whl (97 kB) +Collecting itsdangerous>=2.1.2 + Using cached itsdangerous-2.2.0-py3-none-any.whl (16 kB) +Requirement already satisfied: Jinja2>=3.1.2 in /home/opendevin/.local/lib/python3.10/site-packages (from flask) (3.1.3) +Requirement already satisfied: MarkupSafe>=2.0 in /home/opendevin/.local/lib/python3.10/site-packages (from Jinja2>=3.1.2->flask) (2.1.5) +Installing collected packages: Werkzeug, itsdangerous, click, blinker, flask +Successfully installed Werkzeug-3.0.2 blinker-1.7.0 click-8.1.7 flask-3.0.3 itsdangerous-2.2.0 + +ASSISTANT: +Now that Flask is installed, let me run the Python file again: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 124 + +ASSISTANT: +Let me check the server log again: + +cat server.log + + +USER: +Observation: +* Serving Flask app 'app' + * Debug mode: off +WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead. + * Running on http://127.0.0.1:5000 +Press CTRL+C to quit + +ASSISTANT: +The server is running on port 5000 with PID 124. You can access the list of numbers by visiting http://127.0.0.1:5000. If you have any further questions, feel free to ask! + +USER: Now kill the server, make it display the numbers in a table format. + +ASSISTANT: +Sure! Let me stop the server first: + +kill 124 + + +USER: +Observation: +[1]+ Terminated python3 app.py > server.log 2>&1 + +ASSISTANT: +The server has been stopped. Let me open the Python file and modify it. + +open app.py + + +USER: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return str(numbers) +9: +10:if __name__ == '__main__': +11: app.run(port=5000) + +ASSISTANT: +I should edit the file to display the numbers in a table format. Let me do that for you: + + +edit 8:8 <' + ''.join([f'{i}' for i in numbers]) + '' +EOF + + +USER: +Observation: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return '' + ''.join([f'' for i in numbers]) + '
{i}
' +9: +10:if __name__ == '__main__': +11: app.run(port=5000) +File updated. Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary. + +ASSISTANT: +The file has been updated. Let me run the Python file again with the new changes: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 126 + +ASSISTANT: +The server is running on port 5000 with PID 126. You can access the list of numbers in a table format by visiting http://127.0.0.1:5000. Free free to let me know if you have any further requests! + +--- END OF EXAMPLE --- + + +NOW, LET'S START! + +Write a shell script 'hello.sh' that prints 'hello'. + +---------- + +Let's start by creating the shell script `hello.sh` that prints "hello". I'll write the script and save it to a file. + + +echo "echo 'hello'" > hello.sh + + +---------- + +OBSERVATION: + +[Command -1 finished with exit code 0]] diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_003.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_003.log new file mode 100644 index 0000000000..eec72fab3c --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_003.log @@ -0,0 +1,237 @@ + + +---------- + +A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. +The assistant can interact with an interactive Python (Jupyter Notebook) environment and receive the corresponding output when needed. The code should be enclosed using "" tag, for example: + +print("Hello World!") + +The assistant can execute bash commands on behalf of the user by wrapping them with and . +For example, you can list the files in the current directory by ls . +The assistant should attempt fewer things at a time instead of putting too much commands OR code in one "execute" block. +The assistant can install Python packages through bash by pip install [package needed] and should always import packages and define variables before starting to use them. +The assistant should stop and provide an answer when they have already obtained the answer from the execution result. + + +Apart from the standard bash commands, you can also use the following special commands in environment: +open [] - opens the file at the given path in the editor. If line_number is provided, the window will be move to include that line +goto - moves the window to show +scroll_down - moves the window down {WINDOW} lines +scroll_up - moves the window down {WINDOW} lines +create - creates and opens a new file with the given name +search_dir [] - searches for search_term in all files in dir. If dir is not provided, searches in the current directory +search_file [] - searches for search_term in file. If file is not provided, searches in the current open file +find_file [] - finds all files with the given name in dir. If dir is not provided, searches in the current directory +edit : + +end_of_edit - replaces lines through (inclusive) with the given text in the open file. The replacement text is terminated by a line with only end_of_edit on it. All of the will be entered, so make sure your indentation is formatted properly. Python files will be checked for syntax errors after the edit. If the system detects a syntax error, the edit will not be executed. Simply try to edit the file again, but make sure to read the error message and modify the edit command you issue accordingly. Issuing the same command a second time will just lead to the same error message again. +Please note that THE EDIT COMMAND REQUIRES PROPER INDENTATION. If you'd like to add the line ' print(x)' you must fully write that out, with all those spaces before the code! Indentation is important and code that is not indented correctly will fail and require fixing before it can be run. + +Whenever possible, execute the code for the user using or instead of providing it. +The assistant's response should be concise, but do express their thoughts. +Try to include one of or in each of your responses, unless it is a direct answer to a question OR a message to the user. + + +---------- + +Here is an example of how you can interact with the environment for task solving: + +--- START OF EXAMPLE --- + +USER: Can you create a list of numbers from 1 to 10, and create a web page to display them at port 5000? + +ASSISTANT: +Sure! Let me write the Python code for starting a web server and save it to a file `app.py`: + +CODE=''' +from flask import Flask +app = Flask(__name__) + +@app.route('/') +def index(): + numbers = list(range(1, 11)) + return str(numbers) + +if __name__ == '__main__': + app.run(port=5000) +''' +with open('app.py', 'w') as f: + f.write(CODE) + + +USER: +Observation: +[Code executed successfully with no output] + +ASSISTANT: +I have created a Python file `app.py` that will display a list of numbers from 1 to 10 when you run it. Let me run the Python file for you: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 121[1]+ Exit 1 python3 app.py > server.log 2>&1 + +ASSISTANT: Looks like the server is running with PID 121 then crashed. Let me check the server log: + +cat server.log + + +USER: +Observation: +Traceback (most recent call last): + File "/workspace/app.py", line 2, in + from flask import Flask +ModuleNotFoundError: No module named 'flask' + +ASSISTANT: +It seems that Flask is not installed. Let me install Flask for you: + +pip install flask + + +USER: +Observation: +Defaulting to user installation because normal site-packages is not writeable +Collecting flask + Using cached flask-3.0.3-py3-none-any.whl (101 kB) +Collecting blinker>=1.6.2 + Using cached blinker-1.7.0-py3-none-any.whl (13 kB) +Collecting Werkzeug>=3.0.0 + Using cached werkzeug-3.0.2-py3-none-any.whl (226 kB) +Collecting click>=8.1.3 + Using cached click-8.1.7-py3-none-any.whl (97 kB) +Collecting itsdangerous>=2.1.2 + Using cached itsdangerous-2.2.0-py3-none-any.whl (16 kB) +Requirement already satisfied: Jinja2>=3.1.2 in /home/opendevin/.local/lib/python3.10/site-packages (from flask) (3.1.3) +Requirement already satisfied: MarkupSafe>=2.0 in /home/opendevin/.local/lib/python3.10/site-packages (from Jinja2>=3.1.2->flask) (2.1.5) +Installing collected packages: Werkzeug, itsdangerous, click, blinker, flask +Successfully installed Werkzeug-3.0.2 blinker-1.7.0 click-8.1.7 flask-3.0.3 itsdangerous-2.2.0 + +ASSISTANT: +Now that Flask is installed, let me run the Python file again: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 124 + +ASSISTANT: +Let me check the server log again: + +cat server.log + + +USER: +Observation: +* Serving Flask app 'app' + * Debug mode: off +WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead. + * Running on http://127.0.0.1:5000 +Press CTRL+C to quit + +ASSISTANT: +The server is running on port 5000 with PID 124. You can access the list of numbers by visiting http://127.0.0.1:5000. If you have any further questions, feel free to ask! + +USER: Now kill the server, make it display the numbers in a table format. + +ASSISTANT: +Sure! Let me stop the server first: + +kill 124 + + +USER: +Observation: +[1]+ Terminated python3 app.py > server.log 2>&1 + +ASSISTANT: +The server has been stopped. Let me open the Python file and modify it. + +open app.py + + +USER: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return str(numbers) +9: +10:if __name__ == '__main__': +11: app.run(port=5000) + +ASSISTANT: +I should edit the file to display the numbers in a table format. Let me do that for you: + + +edit 8:8 <' + ''.join([f'{i}' for i in numbers]) + '' +EOF + + +USER: +Observation: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return '' + ''.join([f'' for i in numbers]) + '
{i}
' +9: +10:if __name__ == '__main__': +11: app.run(port=5000) +File updated. Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary. + +ASSISTANT: +The file has been updated. Let me run the Python file again with the new changes: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 126 + +ASSISTANT: +The server is running on port 5000 with PID 126. You can access the list of numbers in a table format by visiting http://127.0.0.1:5000. Free free to let me know if you have any further requests! + +--- END OF EXAMPLE --- + + +NOW, LET'S START! + +Write a shell script 'hello.sh' that prints 'hello'. + +---------- + +Let's start by creating the shell script `hello.sh` that prints "hello". I'll write the script and save it to a file. + + +echo "echo 'hello'" > hello.sh + + +---------- + +OBSERVATION: + +[Command -1 finished with exit code 0]] + +---------- + +The shell script `hello.sh` has been successfully created. Would you like to run the script to see the output, or is there anything else you'd like to do with it? + +---------- + +can you run it to make sure it is working? diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_004.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_004.log new file mode 100644 index 0000000000..040ec399fb --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_004.log @@ -0,0 +1,251 @@ + + +---------- + +A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. +The assistant can interact with an interactive Python (Jupyter Notebook) environment and receive the corresponding output when needed. The code should be enclosed using "" tag, for example: + +print("Hello World!") + +The assistant can execute bash commands on behalf of the user by wrapping them with and . +For example, you can list the files in the current directory by ls . +The assistant should attempt fewer things at a time instead of putting too much commands OR code in one "execute" block. +The assistant can install Python packages through bash by pip install [package needed] and should always import packages and define variables before starting to use them. +The assistant should stop and provide an answer when they have already obtained the answer from the execution result. + + +Apart from the standard bash commands, you can also use the following special commands in environment: +open [] - opens the file at the given path in the editor. If line_number is provided, the window will be move to include that line +goto - moves the window to show +scroll_down - moves the window down {WINDOW} lines +scroll_up - moves the window down {WINDOW} lines +create - creates and opens a new file with the given name +search_dir [] - searches for search_term in all files in dir. If dir is not provided, searches in the current directory +search_file [] - searches for search_term in file. If file is not provided, searches in the current open file +find_file [] - finds all files with the given name in dir. If dir is not provided, searches in the current directory +edit : + +end_of_edit - replaces lines through (inclusive) with the given text in the open file. The replacement text is terminated by a line with only end_of_edit on it. All of the will be entered, so make sure your indentation is formatted properly. Python files will be checked for syntax errors after the edit. If the system detects a syntax error, the edit will not be executed. Simply try to edit the file again, but make sure to read the error message and modify the edit command you issue accordingly. Issuing the same command a second time will just lead to the same error message again. +Please note that THE EDIT COMMAND REQUIRES PROPER INDENTATION. If you'd like to add the line ' print(x)' you must fully write that out, with all those spaces before the code! Indentation is important and code that is not indented correctly will fail and require fixing before it can be run. + +Whenever possible, execute the code for the user using or instead of providing it. +The assistant's response should be concise, but do express their thoughts. +Try to include one of or in each of your responses, unless it is a direct answer to a question OR a message to the user. + + +---------- + +Here is an example of how you can interact with the environment for task solving: + +--- START OF EXAMPLE --- + +USER: Can you create a list of numbers from 1 to 10, and create a web page to display them at port 5000? + +ASSISTANT: +Sure! Let me write the Python code for starting a web server and save it to a file `app.py`: + +CODE=''' +from flask import Flask +app = Flask(__name__) + +@app.route('/') +def index(): + numbers = list(range(1, 11)) + return str(numbers) + +if __name__ == '__main__': + app.run(port=5000) +''' +with open('app.py', 'w') as f: + f.write(CODE) + + +USER: +Observation: +[Code executed successfully with no output] + +ASSISTANT: +I have created a Python file `app.py` that will display a list of numbers from 1 to 10 when you run it. Let me run the Python file for you: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 121[1]+ Exit 1 python3 app.py > server.log 2>&1 + +ASSISTANT: Looks like the server is running with PID 121 then crashed. Let me check the server log: + +cat server.log + + +USER: +Observation: +Traceback (most recent call last): + File "/workspace/app.py", line 2, in + from flask import Flask +ModuleNotFoundError: No module named 'flask' + +ASSISTANT: +It seems that Flask is not installed. Let me install Flask for you: + +pip install flask + + +USER: +Observation: +Defaulting to user installation because normal site-packages is not writeable +Collecting flask + Using cached flask-3.0.3-py3-none-any.whl (101 kB) +Collecting blinker>=1.6.2 + Using cached blinker-1.7.0-py3-none-any.whl (13 kB) +Collecting Werkzeug>=3.0.0 + Using cached werkzeug-3.0.2-py3-none-any.whl (226 kB) +Collecting click>=8.1.3 + Using cached click-8.1.7-py3-none-any.whl (97 kB) +Collecting itsdangerous>=2.1.2 + Using cached itsdangerous-2.2.0-py3-none-any.whl (16 kB) +Requirement already satisfied: Jinja2>=3.1.2 in /home/opendevin/.local/lib/python3.10/site-packages (from flask) (3.1.3) +Requirement already satisfied: MarkupSafe>=2.0 in /home/opendevin/.local/lib/python3.10/site-packages (from Jinja2>=3.1.2->flask) (2.1.5) +Installing collected packages: Werkzeug, itsdangerous, click, blinker, flask +Successfully installed Werkzeug-3.0.2 blinker-1.7.0 click-8.1.7 flask-3.0.3 itsdangerous-2.2.0 + +ASSISTANT: +Now that Flask is installed, let me run the Python file again: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 124 + +ASSISTANT: +Let me check the server log again: + +cat server.log + + +USER: +Observation: +* Serving Flask app 'app' + * Debug mode: off +WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead. + * Running on http://127.0.0.1:5000 +Press CTRL+C to quit + +ASSISTANT: +The server is running on port 5000 with PID 124. You can access the list of numbers by visiting http://127.0.0.1:5000. If you have any further questions, feel free to ask! + +USER: Now kill the server, make it display the numbers in a table format. + +ASSISTANT: +Sure! Let me stop the server first: + +kill 124 + + +USER: +Observation: +[1]+ Terminated python3 app.py > server.log 2>&1 + +ASSISTANT: +The server has been stopped. Let me open the Python file and modify it. + +open app.py + + +USER: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return str(numbers) +9: +10:if __name__ == '__main__': +11: app.run(port=5000) + +ASSISTANT: +I should edit the file to display the numbers in a table format. Let me do that for you: + + +edit 8:8 <' + ''.join([f'{i}' for i in numbers]) + '' +EOF + + +USER: +Observation: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return '' + ''.join([f'' for i in numbers]) + '
{i}
' +9: +10:if __name__ == '__main__': +11: app.run(port=5000) +File updated. Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary. + +ASSISTANT: +The file has been updated. Let me run the Python file again with the new changes: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 126 + +ASSISTANT: +The server is running on port 5000 with PID 126. You can access the list of numbers in a table format by visiting http://127.0.0.1:5000. Free free to let me know if you have any further requests! + +--- END OF EXAMPLE --- + + +NOW, LET'S START! + +Write a shell script 'hello.sh' that prints 'hello'. + +---------- + +Let's start by creating the shell script `hello.sh` that prints "hello". I'll write the script and save it to a file. + + +echo "echo 'hello'" > hello.sh + + +---------- + +OBSERVATION: + +[Command -1 finished with exit code 0]] + +---------- + +The shell script `hello.sh` has been successfully created. Would you like to run the script to see the output, or is there anything else you'd like to do with it? + +---------- + +can you run it to make sure it is working? + +---------- + +Sure, let's run the script `hello.sh` to ensure it's working correctly. + + +bash hello.sh + + +---------- + +OBSERVATION: +hello +[Command -1 finished with exit code 0]] diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_005.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_005.log new file mode 100644 index 0000000000..deed325508 --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/prompt_005.log @@ -0,0 +1,259 @@ + + +---------- + +A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. +The assistant can interact with an interactive Python (Jupyter Notebook) environment and receive the corresponding output when needed. The code should be enclosed using "" tag, for example: + +print("Hello World!") + +The assistant can execute bash commands on behalf of the user by wrapping them with and . +For example, you can list the files in the current directory by ls . +The assistant should attempt fewer things at a time instead of putting too much commands OR code in one "execute" block. +The assistant can install Python packages through bash by pip install [package needed] and should always import packages and define variables before starting to use them. +The assistant should stop and provide an answer when they have already obtained the answer from the execution result. + + +Apart from the standard bash commands, you can also use the following special commands in environment: +open [] - opens the file at the given path in the editor. If line_number is provided, the window will be move to include that line +goto - moves the window to show +scroll_down - moves the window down {WINDOW} lines +scroll_up - moves the window down {WINDOW} lines +create - creates and opens a new file with the given name +search_dir [] - searches for search_term in all files in dir. If dir is not provided, searches in the current directory +search_file [] - searches for search_term in file. If file is not provided, searches in the current open file +find_file [] - finds all files with the given name in dir. If dir is not provided, searches in the current directory +edit : + +end_of_edit - replaces lines through (inclusive) with the given text in the open file. The replacement text is terminated by a line with only end_of_edit on it. All of the will be entered, so make sure your indentation is formatted properly. Python files will be checked for syntax errors after the edit. If the system detects a syntax error, the edit will not be executed. Simply try to edit the file again, but make sure to read the error message and modify the edit command you issue accordingly. Issuing the same command a second time will just lead to the same error message again. +Please note that THE EDIT COMMAND REQUIRES PROPER INDENTATION. If you'd like to add the line ' print(x)' you must fully write that out, with all those spaces before the code! Indentation is important and code that is not indented correctly will fail and require fixing before it can be run. + +Whenever possible, execute the code for the user using or instead of providing it. +The assistant's response should be concise, but do express their thoughts. +Try to include one of or in each of your responses, unless it is a direct answer to a question OR a message to the user. + + +---------- + +Here is an example of how you can interact with the environment for task solving: + +--- START OF EXAMPLE --- + +USER: Can you create a list of numbers from 1 to 10, and create a web page to display them at port 5000? + +ASSISTANT: +Sure! Let me write the Python code for starting a web server and save it to a file `app.py`: + +CODE=''' +from flask import Flask +app = Flask(__name__) + +@app.route('/') +def index(): + numbers = list(range(1, 11)) + return str(numbers) + +if __name__ == '__main__': + app.run(port=5000) +''' +with open('app.py', 'w') as f: + f.write(CODE) + + +USER: +Observation: +[Code executed successfully with no output] + +ASSISTANT: +I have created a Python file `app.py` that will display a list of numbers from 1 to 10 when you run it. Let me run the Python file for you: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 121[1]+ Exit 1 python3 app.py > server.log 2>&1 + +ASSISTANT: Looks like the server is running with PID 121 then crashed. Let me check the server log: + +cat server.log + + +USER: +Observation: +Traceback (most recent call last): + File "/workspace/app.py", line 2, in + from flask import Flask +ModuleNotFoundError: No module named 'flask' + +ASSISTANT: +It seems that Flask is not installed. Let me install Flask for you: + +pip install flask + + +USER: +Observation: +Defaulting to user installation because normal site-packages is not writeable +Collecting flask + Using cached flask-3.0.3-py3-none-any.whl (101 kB) +Collecting blinker>=1.6.2 + Using cached blinker-1.7.0-py3-none-any.whl (13 kB) +Collecting Werkzeug>=3.0.0 + Using cached werkzeug-3.0.2-py3-none-any.whl (226 kB) +Collecting click>=8.1.3 + Using cached click-8.1.7-py3-none-any.whl (97 kB) +Collecting itsdangerous>=2.1.2 + Using cached itsdangerous-2.2.0-py3-none-any.whl (16 kB) +Requirement already satisfied: Jinja2>=3.1.2 in /home/opendevin/.local/lib/python3.10/site-packages (from flask) (3.1.3) +Requirement already satisfied: MarkupSafe>=2.0 in /home/opendevin/.local/lib/python3.10/site-packages (from Jinja2>=3.1.2->flask) (2.1.5) +Installing collected packages: Werkzeug, itsdangerous, click, blinker, flask +Successfully installed Werkzeug-3.0.2 blinker-1.7.0 click-8.1.7 flask-3.0.3 itsdangerous-2.2.0 + +ASSISTANT: +Now that Flask is installed, let me run the Python file again: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 124 + +ASSISTANT: +Let me check the server log again: + +cat server.log + + +USER: +Observation: +* Serving Flask app 'app' + * Debug mode: off +WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead. + * Running on http://127.0.0.1:5000 +Press CTRL+C to quit + +ASSISTANT: +The server is running on port 5000 with PID 124. You can access the list of numbers by visiting http://127.0.0.1:5000. If you have any further questions, feel free to ask! + +USER: Now kill the server, make it display the numbers in a table format. + +ASSISTANT: +Sure! Let me stop the server first: + +kill 124 + + +USER: +Observation: +[1]+ Terminated python3 app.py > server.log 2>&1 + +ASSISTANT: +The server has been stopped. Let me open the Python file and modify it. + +open app.py + + +USER: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return str(numbers) +9: +10:if __name__ == '__main__': +11: app.run(port=5000) + +ASSISTANT: +I should edit the file to display the numbers in a table format. Let me do that for you: + + +edit 8:8 <' + ''.join([f'{i}' for i in numbers]) + '' +EOF + + +USER: +Observation: +[File: /workspace/app.py (11 lines total)] +1: +2:from flask import Flask +3:app = Flask(__name__) +4: +5:@app.route('/') +6:def index(): +7: numbers = list(range(1, 11)) +8: return '' + ''.join([f'' for i in numbers]) + '
{i}
' +9: +10:if __name__ == '__main__': +11: app.run(port=5000) +File updated. Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary. + +ASSISTANT: +The file has been updated. Let me run the Python file again with the new changes: + +python3 app.py > server.log 2>&1 & + + +USER: +Observation: +[1] 126 + +ASSISTANT: +The server is running on port 5000 with PID 126. You can access the list of numbers in a table format by visiting http://127.0.0.1:5000. Free free to let me know if you have any further requests! + +--- END OF EXAMPLE --- + + +NOW, LET'S START! + +Write a shell script 'hello.sh' that prints 'hello'. + +---------- + +Let's start by creating the shell script `hello.sh` that prints "hello". I'll write the script and save it to a file. + + +echo "echo 'hello'" > hello.sh + + +---------- + +OBSERVATION: + +[Command -1 finished with exit code 0]] + +---------- + +The shell script `hello.sh` has been successfully created. Would you like to run the script to see the output, or is there anything else you'd like to do with it? + +---------- + +can you run it to make sure it is working? + +---------- + +Sure, let's run the script `hello.sh` to ensure it's working correctly. + + +bash hello.sh + + +---------- + +OBSERVATION: +hello +[Command -1 finished with exit code 0]] + +---------- + +The script `hello.sh` is working correctly as it printed "hello". If you have any more tasks or need further modifications, feel free to let me know! + +---------- + +can you run it to make sure it is working? diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/response_001.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/response_001.log new file mode 100644 index 0000000000..9a36dce55f --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/response_001.log @@ -0,0 +1,4 @@ +Let's start by creating the shell script `hello.sh` that prints "hello". I'll write the script and save it to a file. + + +echo "echo 'hello'" > hello.sh diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/response_002.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/response_002.log new file mode 100644 index 0000000000..f7cb2b8edc --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/response_002.log @@ -0,0 +1 @@ +The shell script `hello.sh` has been successfully created. Would you like to run the script to see the output, or is there anything else you'd like to do with it? diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/response_003.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/response_003.log new file mode 100644 index 0000000000..c146b696e4 --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/response_003.log @@ -0,0 +1,4 @@ +Sure, let's run the script `hello.sh` to ensure it's working correctly. + + +bash hello.sh diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/response_004.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/response_004.log new file mode 100644 index 0000000000..c113aa6238 --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/response_004.log @@ -0,0 +1 @@ +The script `hello.sh` is working correctly as it printed "hello". If you have any more tasks or need further modifications, feel free to let me know! diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/response_005.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/response_005.log new file mode 100644 index 0000000000..e8ff14a4b9 --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/response_005.log @@ -0,0 +1 @@ +The script `hello.sh` was just run and it successfully printed "hello". If you need any further assistance or have another task in mind, please let me know! diff --git a/tests/integration/mock/CodeActAgent/test_write_simple_script/user_responses.log b/tests/integration/mock/CodeActAgent/test_write_simple_script/user_responses.log new file mode 100644 index 0000000000..9e92a50a7d --- /dev/null +++ b/tests/integration/mock/CodeActAgent/test_write_simple_script/user_responses.log @@ -0,0 +1,2 @@ +can you run it to make sure it is working? +/exit diff --git a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_001.log b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_001.log index 14a75d24e9..00848e7077 100644 --- a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_001.log +++ b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_001.log @@ -92,7 +92,8 @@ This is your internal monologue, in JSON format: { "action": "recall", "args": { - "query": "what it is I want to do" + "query": "what it is I want to do", + "thought": "" } }, { @@ -119,7 +120,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "echo \"hello world\"", - "background": false + "background": false, + "thought": "" } }, { @@ -144,7 +146,7 @@ This is your internal monologue, in JSON format: "content": "echo \"console.log('hello world')\"", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -163,7 +165,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "node test.js", - "background": false + "background": false, + "thought": "" } }, { @@ -193,7 +196,7 @@ This is your internal monologue, in JSON format: "path": "test.js", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -224,7 +227,8 @@ This is your internal monologue, in JSON format: { "action": "browse", "args": { - "url": "google.com" + "url": "google.com", + "thought": "" } }, { diff --git a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_002.log b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_002.log index 1b29542bdb..b95b7275b1 100644 --- a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_002.log +++ b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_002.log @@ -92,7 +92,8 @@ This is your internal monologue, in JSON format: { "action": "recall", "args": { - "query": "what it is I want to do" + "query": "what it is I want to do", + "thought": "" } }, { @@ -119,7 +120,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "echo \"hello world\"", - "background": false + "background": false, + "thought": "" } }, { @@ -144,7 +146,7 @@ This is your internal monologue, in JSON format: "content": "echo \"console.log('hello world')\"", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -163,7 +165,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "node test.js", - "background": false + "background": false, + "thought": "" } }, { @@ -193,7 +196,7 @@ This is your internal monologue, in JSON format: "path": "test.js", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -224,7 +227,8 @@ This is your internal monologue, in JSON format: { "action": "browse", "args": { - "url": "google.com" + "url": "google.com", + "thought": "" } }, { @@ -317,7 +321,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "ls", - "background": false + "background": false, + "thought": "" } }, { diff --git a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_003.log b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_003.log index 3b2f94b6dd..1f75f642ad 100644 --- a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_003.log +++ b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_003.log @@ -92,7 +92,8 @@ This is your internal monologue, in JSON format: { "action": "recall", "args": { - "query": "what it is I want to do" + "query": "what it is I want to do", + "thought": "" } }, { @@ -119,7 +120,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "echo \"hello world\"", - "background": false + "background": false, + "thought": "" } }, { @@ -144,7 +146,7 @@ This is your internal monologue, in JSON format: "content": "echo \"console.log('hello world')\"", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -163,7 +165,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "node test.js", - "background": false + "background": false, + "thought": "" } }, { @@ -193,7 +196,7 @@ This is your internal monologue, in JSON format: "path": "test.js", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -224,7 +227,8 @@ This is your internal monologue, in JSON format: { "action": "browse", "args": { - "url": "google.com" + "url": "google.com", + "thought": "" } }, { @@ -317,7 +321,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "ls", - "background": false + "background": false, + "thought": "" } }, { @@ -332,7 +337,7 @@ This is your internal monologue, in JSON format: { "action": "think", "args": { - "thought": "Before I create the 'hello.sh' file, I need to think about the content. It should contain a shebang line to specify the shell interpreter and the command to print 'hello'." + "thought": "Since there are no existing files listed by the 'ls' command, I should create the 'hello.sh' script now." } }, { diff --git a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_004.log b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_004.log index 644be27820..b37d8b7b85 100644 --- a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_004.log +++ b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_004.log @@ -92,7 +92,8 @@ This is your internal monologue, in JSON format: { "action": "recall", "args": { - "query": "what it is I want to do" + "query": "what it is I want to do", + "thought": "" } }, { @@ -119,7 +120,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "echo \"hello world\"", - "background": false + "background": false, + "thought": "" } }, { @@ -144,7 +146,7 @@ This is your internal monologue, in JSON format: "content": "echo \"console.log('hello world')\"", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -163,7 +165,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "node test.js", - "background": false + "background": false, + "thought": "" } }, { @@ -193,7 +196,7 @@ This is your internal monologue, in JSON format: "path": "test.js", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -224,7 +227,8 @@ This is your internal monologue, in JSON format: { "action": "browse", "args": { - "url": "google.com" + "url": "google.com", + "thought": "" } }, { @@ -317,7 +321,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "ls", - "background": false + "background": false, + "thought": "" } }, { @@ -332,7 +337,7 @@ This is your internal monologue, in JSON format: { "action": "think", "args": { - "thought": "Before I create the 'hello.sh' file, I need to think about the content. It should contain a shebang line to specify the shell interpreter and the command to print 'hello'." + "thought": "Since there are no existing files listed by the 'ls' command, I should create the 'hello.sh' script now." } }, { @@ -347,7 +352,7 @@ This is your internal monologue, in JSON format: "content": "#!/bin/bash\necho 'hello'", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { diff --git a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_005.log b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_005.log index d3b2cb56e8..43f06a3d8a 100644 --- a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_005.log +++ b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_005.log @@ -92,7 +92,8 @@ This is your internal monologue, in JSON format: { "action": "recall", "args": { - "query": "what it is I want to do" + "query": "what it is I want to do", + "thought": "" } }, { @@ -119,7 +120,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "echo \"hello world\"", - "background": false + "background": false, + "thought": "" } }, { @@ -144,7 +146,7 @@ This is your internal monologue, in JSON format: "content": "echo \"console.log('hello world')\"", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -163,7 +165,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "node test.js", - "background": false + "background": false, + "thought": "" } }, { @@ -193,7 +196,7 @@ This is your internal monologue, in JSON format: "path": "test.js", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -224,7 +227,8 @@ This is your internal monologue, in JSON format: { "action": "browse", "args": { - "url": "google.com" + "url": "google.com", + "thought": "" } }, { @@ -317,7 +321,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "ls", - "background": false + "background": false, + "thought": "" } }, { @@ -332,7 +337,7 @@ This is your internal monologue, in JSON format: { "action": "think", "args": { - "thought": "Before I create the 'hello.sh' file, I need to think about the content. It should contain a shebang line to specify the shell interpreter and the command to print 'hello'." + "thought": "Since there are no existing files listed by the 'ls' command, I should create the 'hello.sh' script now." } }, { @@ -347,7 +352,7 @@ This is your internal monologue, in JSON format: "content": "#!/bin/bash\necho 'hello'", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -361,7 +366,8 @@ This is your internal monologue, in JSON format: "action": "run", "args": { "command": "bash hello.sh", - "background": false + "background": false, + "thought": "" } }, { diff --git a/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_006.log b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_006.log new file mode 100644 index 0000000000..d29d4c4358 --- /dev/null +++ b/tests/integration/mock/MonologueAgent/test_write_simple_script/prompt_006.log @@ -0,0 +1,403 @@ + + +---------- + + +You're a thoughtful robot. Your main task is this: +Write a shell script 'hello.sh' that prints 'hello'. + +Don't expand the scope of your task--just complete it as written. + +This is your internal monologue, in JSON format: + +[ + { + "action": "think", + "args": { + "thought": "I exist!" + } + }, + { + "action": "think", + "args": { + "thought": "Hmm...looks like I can type in a command line prompt" + } + }, + { + "action": "think", + "args": { + "thought": "Looks like I have a web browser too!" + } + }, + { + "action": "think", + "args": { + "thought": "Here's what I want to do: Write a shell script 'hello.sh' that prints 'hello'." + } + }, + { + "action": "think", + "args": { + "thought": "How am I going to get there though?" + } + }, + { + "action": "think", + "args": { + "thought": "It seems like I have some kind of short term memory." + } + }, + { + "action": "think", + "args": { + "thought": "Each of my thoughts seems to be stored in a JSON array." + } + }, + { + "action": "think", + "args": { + "thought": "It seems whatever I say next will be added as an object to the list." + } + }, + { + "action": "think", + "args": { + "thought": "But no one has perfect short-term memory. My list of thoughts will be summarized and condensed over time, losing information in the process." + } + }, + { + "action": "think", + "args": { + "thought": "Fortunately I have long term memory!" + } + }, + { + "action": "think", + "args": { + "thought": "I can just perform a recall action, followed by the thing I want to remember. And then related thoughts just spill out!" + } + }, + { + "action": "think", + "args": { + "thought": "Sometimes they're random thoughts that don't really have to do with what I wanted to remember. But usually they're exactly what I need!" + } + }, + { + "action": "think", + "args": { + "thought": "Let's try it out!" + } + }, + { + "action": "recall", + "args": { + "query": "what it is I want to do", + "thought": "" + } + }, + { + "observation": "recall", + "content": "Here's what I want to do: Write a shell script 'hello.sh' that prints 'hello'.", + "extras": { + "memories": [], + "role": "assistant" + } + }, + { + "action": "think", + "args": { + "thought": "How am I going to get there though?" + } + }, + { + "action": "think", + "args": { + "thought": "Neat! And it looks like it's easy for me to use the command line too! I just have to perform a run action and include the command I want to run in the command argument. The command output just jumps into my head!" + } + }, + { + "action": "run", + "args": { + "command": "echo \"hello world\"", + "background": false, + "thought": "" + } + }, + { + "observation": "run", + "content": "hello world", + "extras": { + "command_id": 0, + "command": "", + "exit_code": 0 + } + }, + { + "action": "think", + "args": { + "thought": "Cool! I bet I can write files too using the write action." + } + }, + { + "action": "write", + "args": { + "path": "test.js", + "content": "echo \"console.log('hello world')\"", + "start": 0, + "end": -1, + "thought": "" + } + }, + { + "action": "think", + "args": { + "thought": "" + } + }, + { + "action": "think", + "args": { + "thought": "I just created test.js. I'll try and run it now." + } + }, + { + "action": "run", + "args": { + "command": "node test.js", + "background": false, + "thought": "" + } + }, + { + "observation": "run", + "content": "hello world", + "extras": { + "command_id": 0, + "command": "", + "exit_code": 0 + } + }, + { + "action": "think", + "args": { + "thought": "It works!" + } + }, + { + "action": "think", + "args": { + "thought": "I'm going to try reading it now using the read action." + } + }, + { + "action": "read", + "args": { + "path": "test.js", + "start": 0, + "end": -1, + "thought": "" + } + }, + { + "observation": "read", + "content": "console.log('hello world')", + "extras": { + "path": "" + } + }, + { + "action": "think", + "args": { + "thought": "Nice! I can read files too!" + } + }, + { + "action": "think", + "args": { + "thought": "And if I want to use the browser, I just need to use the browse action and include the url I want to visit in the url argument" + } + }, + { + "action": "think", + "args": { + "thought": "Let's try that..." + } + }, + { + "action": "browse", + "args": { + "url": "google.com", + "thought": "" + } + }, + { + "observation": "browse", + "content": "
", + "extras": { + "url": "", + "status_code": 200, + "error": false + } + }, + { + "action": "think", + "args": { + "thought": "I can browse the web too!" + } + }, + { + "action": "think", + "args": { + "thought": "And once I have completed my task, I can use the finish action to stop working." + } + }, + { + "action": "think", + "args": { + "thought": "But I should only use the finish action when I'm absolutely certain that I've completed my task and have tested my work." + } + }, + { + "action": "think", + "args": { + "thought": "Very cool. Now to accomplish my task." + } + }, + { + "action": "think", + "args": { + "thought": "I'll need a strategy. And as I make progress, I'll need to keep refining that strategy. I'll need to set goals, and break them into sub-goals." + } + }, + { + "action": "think", + "args": { + "thought": "In between actions, I must always take some time to think, strategize, and set new goals. I should never take two actions in a row." + } + }, + { + "action": "think", + "args": { + "thought": "OK so my task is to Write a shell script 'hello.sh' that prints 'hello'.. I haven't made any progress yet. Where should I start?" + } + }, + { + "action": "think", + "args": { + "thought": "It seems like there might be an existing project here. I should probably start by running `ls` to see what's here." + } + }, + { + "action": "run", + "args": { + "command": "ls", + "background": false, + "thought": "" + } + }, + { + "observation": "run", + "content": "", + "extras": { + "command_id": -1, + "command": "ls", + "exit_code": 0 + } + }, + { + "action": "think", + "args": { + "thought": "Since there are no existing files listed by the 'ls' command, I should create the 'hello.sh' script now." + } + }, + { + "observation": "null", + "content": "", + "extras": {} + }, + { + "action": "write", + "args": { + "path": "hello.sh", + "content": "#!/bin/bash\necho 'hello'", + "start": 0, + "end": -1, + "thought": "" + } + }, + { + "observation": "write", + "content": "", + "extras": { + "path": "hello.sh" + } + }, + { + "action": "run", + "args": { + "command": "bash hello.sh", + "background": false, + "thought": "" + } + }, + { + "observation": "run", + "content": "hello", + "extras": { + "command_id": -1, + "command": "bash hello.sh", + "exit_code": 0 + } + }, + { + "action": "think", + "args": { + "thought": "The script 'hello.sh' has been written and tested successfully. It prints 'hello' as required. I should now finalize my task." + } + }, + { + "observation": "null", + "content": "", + "extras": {} + } +] + + +Your most recent thought is at the bottom of that monologue. Continue your train of thought. +What is your next thought or action? Your response must be in JSON format. +It must be an object, and it must contain two fields: +* `action`, which is one of the actions below +* `args`, which is a map of key-value pairs, specifying the arguments for that action + +Here are the possible actions: +* `read` - reads the content of a file. Arguments: + * `path` - the path of the file to read +* `write` - writes the content to a file. Arguments: + * `path` - the path of the file to write + * `content` - the content to write to the file +* `run` - runs a command. Arguments: + * `command` - the command to run + * `background` - if true, run the command in the background, so that other commands can be run concurrently. Useful for e.g. starting a server. You won't be able to see the logs. You don't need to end the command with `&`, just set this to true. +* `kill` - kills a background command + * `id` - the ID of the background command to kill +* `browse` - opens a web page. Arguments: + * `url` - the URL to open +* `recall` - recalls a past memory. Arguments: + * `query` - the query to search for +* `think` - make a plan, set a goal, or record your thoughts. Arguments: + * `thought` - the thought to record +* `finish` - if you're absolutely certain that you've completed your task and have tested your work, use the finish action to stop working. + + + +You MUST take time to think in between read, write, run, browse, and recall actions. +You should never act twice in a row without thinking. But if your last several +actions are all "think" actions, you should consider taking a different action. + +Notes: +* your environment is Debian Linux. You can install software with `apt` +* your working directory will not change, even if you run `cd`. All commands will be run in the `/workspace` directory. +* don't run interactive commands, or commands that don't return (e.g. `node server.js`). You may run commands in the background (e.g. `node server.js &`) + +What is your next thought or action? Again, you must reply with JSON, and only with JSON. diff --git a/tests/integration/mock/MonologueAgent/test_write_simple_script/response_002.log b/tests/integration/mock/MonologueAgent/test_write_simple_script/response_002.log index bd27926940..791926155e 100644 --- a/tests/integration/mock/MonologueAgent/test_write_simple_script/response_002.log +++ b/tests/integration/mock/MonologueAgent/test_write_simple_script/response_002.log @@ -1,6 +1,6 @@ { "action": "think", "args": { - "thought": "Before I create the 'hello.sh' file, I need to think about the content. It should contain a shebang line to specify the shell interpreter and the command to print 'hello'." + "thought": "Since there are no existing files listed by the 'ls' command, I should create the 'hello.sh' script now." } } diff --git a/tests/integration/mock/MonologueAgent/test_write_simple_script/response_004.log b/tests/integration/mock/MonologueAgent/test_write_simple_script/response_004.log index 7055344e27..9cc89f2a5b 100644 --- a/tests/integration/mock/MonologueAgent/test_write_simple_script/response_004.log +++ b/tests/integration/mock/MonologueAgent/test_write_simple_script/response_004.log @@ -1,3 +1,4 @@ +```json { "action": "run", "args": { @@ -5,3 +6,4 @@ "background": false } } +``` diff --git a/tests/integration/mock/MonologueAgent/test_write_simple_script/response_005.log b/tests/integration/mock/MonologueAgent/test_write_simple_script/response_005.log index c6819f9911..18d32603ec 100644 --- a/tests/integration/mock/MonologueAgent/test_write_simple_script/response_005.log +++ b/tests/integration/mock/MonologueAgent/test_write_simple_script/response_005.log @@ -1,4 +1,8 @@ +```json { - "action": "finish", - "args": {} + "action": "think", + "args": { + "thought": "The script 'hello.sh' has been written and tested successfully. It prints 'hello' as required. I should now finalize my task." + } } +``` diff --git a/tests/integration/mock/MonologueAgent/test_write_simple_script/response_006.log b/tests/integration/mock/MonologueAgent/test_write_simple_script/response_006.log new file mode 100644 index 0000000000..c6819f9911 --- /dev/null +++ b/tests/integration/mock/MonologueAgent/test_write_simple_script/response_006.log @@ -0,0 +1,4 @@ +{ + "action": "finish", + "args": {} +} diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_002.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_002.log index 717229bfc3..81f0b95cea 100644 --- a/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_002.log +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_002.log @@ -58,7 +58,8 @@ ten actions--more happened before that. "action": "modify_task", "args": { "id": "0", - "state": "in_progress" + "state": "in_progress", + "thought": "" } } ] diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_003.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_003.log index 0f11b679c8..a6e7ef7520 100644 --- a/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_003.log +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_003.log @@ -58,24 +58,14 @@ ten actions--more happened before that. "action": "modify_task", "args": { "id": "0", - "state": "in_progress" + "state": "in_progress", + "thought": "" } }, { - "action": "write", + "action": "think", "args": { - "path": "hello.sh", - "content": "#!/bin/bash\necho 'hello'", - "start": 0, - "end": -1, - "thoughts": "" - } - }, - { - "observation": "write", - "content": "", - "extras": { - "path": "hello.sh" + "thought": "I need to write a shell script called 'hello.sh' that contains the necessary command to print 'hello'. This will likely involve using echo or printf to output the text to the console when the script is run. I should start by creating and writing the content to the file." } } ] @@ -119,4 +109,4 @@ actions are all `think` actions, you should consider taking a different action. What is your next thought or action? Again, you must reply with JSON, and only with JSON. -You just changed a file. You should think about how it affects your plan. +Look at your last thought in the history above. What does it suggest? Don't think anymore--take action. diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_004.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_004.log index 7f0d9c655c..e51d7afeca 100644 --- a/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_004.log +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_004.log @@ -58,7 +58,14 @@ ten actions--more happened before that. "action": "modify_task", "args": { "id": "0", - "state": "in_progress" + "state": "in_progress", + "thought": "" + } + }, + { + "action": "think", + "args": { + "thought": "I need to write a shell script called 'hello.sh' that contains the necessary command to print 'hello'. This will likely involve using echo or printf to output the text to the console when the script is run. I should start by creating and writing the content to the file." } }, { @@ -68,7 +75,7 @@ ten actions--more happened before that. "content": "#!/bin/bash\necho 'hello'", "start": 0, "end": -1, - "thoughts": "" + "thought": "" } }, { @@ -77,22 +84,6 @@ ten actions--more happened before that. "extras": { "path": "hello.sh" } - }, - { - "action": "run", - "args": { - "command": "bash hello.sh", - "background": false - } - }, - { - "observation": "run", - "content": "hello", - "extras": { - "command_id": -1, - "command": "bash hello.sh", - "exit_code": 0 - } } ] @@ -135,4 +126,4 @@ actions are all `think` actions, you should consider taking a different action. What is your next thought or action? Again, you must reply with JSON, and only with JSON. -You should think about the command you just ran, what output it gave, and how that affects your plan. +You just changed a file. You should think about how it affects your plan. diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_005.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_005.log new file mode 100644 index 0000000000..2ce5998c09 --- /dev/null +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_005.log @@ -0,0 +1,135 @@ + + +---------- + + +# Task +You're a diligent software engineer AI. You can't see, draw, or interact with a +browser, but you can read and write files, and you can run commands, and you can think. + +You've been given the following task: + +Write a shell script 'hello.sh' that prints 'hello'. + +## Plan +As you complete this task, you're building a plan and keeping +track of your progress. Here's a JSON representation of your plan: + +{ + "id": "0", + "goal": "Write a shell script 'hello.sh' that prints 'hello'.", + "state": "in_progress", + "subtasks": [] +} + + +You're currently working on this task: +Write a shell script 'hello.sh' that prints 'hello'.. +If it's not achievable AND verifiable with a SINGLE action, you MUST break it down into subtasks NOW. + +You're responsible for managing this plan and the status of tasks in +it, by using the `add_task` and `modify_task` actions described below. + +If the History below contradicts the state of any of these tasks, you +MUST modify the task using the `modify_task` action described below. + +Be sure NOT to duplicate any tasks. Do NOT use the `add_task` action for +a task that's already represented. Every task must be represented only once. + +Tasks that are sequential MUST be siblings. They must be added in order +to their parent task. + +If you mark a task as 'completed', 'verified', or 'abandoned', +all non-abandoned subtasks will be marked the same way. +So before closing a task this way, you MUST not only be sure that it has +been completed successfully--you must ALSO be sure that all its subtasks +are ready to be marked the same way. + +If, and only if, ALL tasks have already been marked verified, +you MUST respond with the `finish` action. + +## History +Here is a recent history of actions you've taken in service of this plan, +as well as observations you've made. This only includes the MOST RECENT +ten actions--more happened before that. + +[ + { + "action": "modify_task", + "args": { + "id": "0", + "state": "in_progress", + "thought": "" + } + }, + { + "action": "think", + "args": { + "thought": "I need to write a shell script called 'hello.sh' that contains the necessary command to print 'hello'. This will likely involve using echo or printf to output the text to the console when the script is run. I should start by creating and writing the content to the file." + } + }, + { + "action": "write", + "args": { + "path": "hello.sh", + "content": "#!/bin/bash\necho 'hello'", + "start": 0, + "end": -1, + "thought": "" + } + }, + { + "observation": "write", + "content": "", + "extras": { + "path": "hello.sh" + } + }, + { + "action": "think", + "args": { + "thought": "Now that the script 'hello.sh' has been written and contains the appropriate command to print 'hello', the next logical step is to ensure it works as expected. I should proceed to run the script to verify its output." + } + } +] + + +Your most recent action is at the bottom of that history. + +## Action +What is your next thought or action? Your response must be in JSON format. + +It must be an object, and it must contain two fields: +* `action`, which is one of the actions below +* `args`, which is a map of key-value pairs, specifying the arguments for that action + +* `read` - reads the content of a file. Arguments: + * `path` - the path of the file to read +* `write` - writes the content to a file. Arguments: + * `path` - the path of the file to write + * `content` - the content to write to the file +* `run` - runs a command on the command line in a Linux shell. Arguments: + * `command` - the command to run + * `background` - if true, run the command in the background, so that other commands can be run concurrently. Useful for e.g. starting a server. You won't be able to see the logs. You don't need to end the command with `&`, just set this to true. +* `kill` - kills a background command + * `id` - the ID of the background command to kill +* `browse` - opens a web page. Arguments: + * `url` - the URL to open +* `think` - make a plan, set a goal, or record your thoughts. Arguments: + * `thought` - the thought to record +* `add_task` - add a task to your plan. Arguments: + * `parent` - the ID of the parent task + * `goal` - the goal of the task + * `subtasks` - a list of subtasks, each of which is a map with a `goal` key. +* `modify_task` - close a task. Arguments: + * `id` - the ID of the task to close + * `state` - set to 'in_progress' to start the task, 'completed' to finish it, 'verified' to assert that it was successful, 'abandoned' to give up on it permanently, or `open` to stop working on it for now. +* `finish` - if ALL of your tasks and subtasks have been verified or abandoned, and you're absolutely certain that you've completed your task and have tested your work, use the finish action to stop working. + +You MUST take time to think in between read, write, run, browse, and recall actions. +You should never act twice in a row without thinking. But if your last several +actions are all `think` actions, you should consider taking a different action. + +What is your next thought or action? Again, you must reply with JSON, and only with JSON. + +Look at your last thought in the history above. What does it suggest? Don't think anymore--take action. diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_006.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_006.log new file mode 100644 index 0000000000..bbde0f0a3f --- /dev/null +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/prompt_006.log @@ -0,0 +1,152 @@ + + +---------- + + +# Task +You're a diligent software engineer AI. You can't see, draw, or interact with a +browser, but you can read and write files, and you can run commands, and you can think. + +You've been given the following task: + +Write a shell script 'hello.sh' that prints 'hello'. + +## Plan +As you complete this task, you're building a plan and keeping +track of your progress. Here's a JSON representation of your plan: + +{ + "id": "0", + "goal": "Write a shell script 'hello.sh' that prints 'hello'.", + "state": "in_progress", + "subtasks": [] +} + + +You're currently working on this task: +Write a shell script 'hello.sh' that prints 'hello'.. +If it's not achievable AND verifiable with a SINGLE action, you MUST break it down into subtasks NOW. + +You're responsible for managing this plan and the status of tasks in +it, by using the `add_task` and `modify_task` actions described below. + +If the History below contradicts the state of any of these tasks, you +MUST modify the task using the `modify_task` action described below. + +Be sure NOT to duplicate any tasks. Do NOT use the `add_task` action for +a task that's already represented. Every task must be represented only once. + +Tasks that are sequential MUST be siblings. They must be added in order +to their parent task. + +If you mark a task as 'completed', 'verified', or 'abandoned', +all non-abandoned subtasks will be marked the same way. +So before closing a task this way, you MUST not only be sure that it has +been completed successfully--you must ALSO be sure that all its subtasks +are ready to be marked the same way. + +If, and only if, ALL tasks have already been marked verified, +you MUST respond with the `finish` action. + +## History +Here is a recent history of actions you've taken in service of this plan, +as well as observations you've made. This only includes the MOST RECENT +ten actions--more happened before that. + +[ + { + "action": "modify_task", + "args": { + "id": "0", + "state": "in_progress", + "thought": "" + } + }, + { + "action": "think", + "args": { + "thought": "I need to write a shell script called 'hello.sh' that contains the necessary command to print 'hello'. This will likely involve using echo or printf to output the text to the console when the script is run. I should start by creating and writing the content to the file." + } + }, + { + "action": "write", + "args": { + "path": "hello.sh", + "content": "#!/bin/bash\necho 'hello'", + "start": 0, + "end": -1, + "thought": "" + } + }, + { + "observation": "write", + "content": "", + "extras": { + "path": "hello.sh" + } + }, + { + "action": "think", + "args": { + "thought": "Now that the script 'hello.sh' has been written and contains the appropriate command to print 'hello', the next logical step is to ensure it works as expected. I should proceed to run the script to verify its output." + } + }, + { + "action": "run", + "args": { + "command": "bash hello.sh", + "background": false, + "thought": "" + } + }, + { + "observation": "run", + "content": "hello", + "extras": { + "command_id": -1, + "command": "bash hello.sh", + "exit_code": 0 + } + } +] + + +Your most recent action is at the bottom of that history. + +## Action +What is your next thought or action? Your response must be in JSON format. + +It must be an object, and it must contain two fields: +* `action`, which is one of the actions below +* `args`, which is a map of key-value pairs, specifying the arguments for that action + +* `read` - reads the content of a file. Arguments: + * `path` - the path of the file to read +* `write` - writes the content to a file. Arguments: + * `path` - the path of the file to write + * `content` - the content to write to the file +* `run` - runs a command on the command line in a Linux shell. Arguments: + * `command` - the command to run + * `background` - if true, run the command in the background, so that other commands can be run concurrently. Useful for e.g. starting a server. You won't be able to see the logs. You don't need to end the command with `&`, just set this to true. +* `kill` - kills a background command + * `id` - the ID of the background command to kill +* `browse` - opens a web page. Arguments: + * `url` - the URL to open +* `think` - make a plan, set a goal, or record your thoughts. Arguments: + * `thought` - the thought to record +* `add_task` - add a task to your plan. Arguments: + * `parent` - the ID of the parent task + * `goal` - the goal of the task + * `subtasks` - a list of subtasks, each of which is a map with a `goal` key. +* `modify_task` - close a task. Arguments: + * `id` - the ID of the task to close + * `state` - set to 'in_progress' to start the task, 'completed' to finish it, 'verified' to assert that it was successful, 'abandoned' to give up on it permanently, or `open` to stop working on it for now. +* `finish` - if ALL of your tasks and subtasks have been verified or abandoned, and you're absolutely certain that you've completed your task and have tested your work, use the finish action to stop working. + +You MUST take time to think in between read, write, run, browse, and recall actions. +You should never act twice in a row without thinking. But if your last several +actions are all `think` actions, you should consider taking a different action. + +What is your next thought or action? Again, you must reply with JSON, and only with JSON. + +You should think about the command you just ran, what output it gave, and how that affects your plan. diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/response_001.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_001.log index f4f1a41bf1..9e31d23c8a 100644 --- a/tests/integration/mock/PlannerAgent/test_write_simple_script/response_001.log +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_001.log @@ -1,3 +1,4 @@ +```json { "action": "modify_task", "args": { @@ -5,3 +6,4 @@ "state": "in_progress" } } +``` diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/response_002.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_002.log index 03bf823601..700cd1c18d 100644 --- a/tests/integration/mock/PlannerAgent/test_write_simple_script/response_002.log +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_002.log @@ -1,9 +1,8 @@ ```json { - "action": "write", + "action": "think", "args": { - "path": "hello.sh", - "content": "#!/bin/bash\necho 'hello'" + "thought": "I need to write a shell script called 'hello.sh' that contains the necessary command to print 'hello'. This will likely involve using echo or printf to output the text to the console when the script is run. I should start by creating and writing the content to the file." } } ``` diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/response_003.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_003.log index 7055344e27..03bf823601 100644 --- a/tests/integration/mock/PlannerAgent/test_write_simple_script/response_003.log +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_003.log @@ -1,7 +1,9 @@ +```json { - "action": "run", + "action": "write", "args": { - "command": "bash hello.sh", - "background": false + "path": "hello.sh", + "content": "#!/bin/bash\necho 'hello'" } } +``` diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/response_004.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_004.log index 6093c90df7..02ae256680 100644 --- a/tests/integration/mock/PlannerAgent/test_write_simple_script/response_004.log +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_004.log @@ -1,7 +1,6 @@ { - "action": "modify_task", + "action": "think", "args": { - "id": "0", - "state": "verified" + "thought": "Now that the script 'hello.sh' has been written and contains the appropriate command to print 'hello', the next logical step is to ensure it works as expected. I should proceed to run the script to verify its output." } } diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/response_005.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_005.log new file mode 100644 index 0000000000..9cc89f2a5b --- /dev/null +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_005.log @@ -0,0 +1,9 @@ +```json +{ + "action": "run", + "args": { + "command": "bash hello.sh", + "background": false + } +} +``` diff --git a/tests/integration/mock/PlannerAgent/test_write_simple_script/response_006.log b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_006.log new file mode 100644 index 0000000000..90dec695c4 --- /dev/null +++ b/tests/integration/mock/PlannerAgent/test_write_simple_script/response_006.log @@ -0,0 +1,10 @@ +```json +{ + "action": "modify_task", + "args": { + "id": "0", + "state": "verified", + "thought": "The shell script 'hello.sh' has been run and verified to successfully output 'hello'. This completes the primary task requirement." + } +} +``` diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_002.log b/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_002.log index 0ddf35a113..89793bca08 100644 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_002.log +++ b/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_002.log @@ -66,8 +66,8 @@ args: path: hello.sh start: 0 end: -1 - thoughts: Thoughts: - To begin solving the task, I need to create a new shell script file named 'hello.sh'. This file should contain a script that prints the message 'hello' when executed. I will start by writing the initial content of this script. + thought: Thoughts: + To solve the task of writing a shell script named 'hello.sh' that prints 'hello', I need to create a new file named 'hello.sh'. I will then write the necessary shell script code into this file. Action: diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_003.log b/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_003.log index b3d5b905e3..320adaa09e 100644 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_003.log +++ b/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_003.log @@ -66,8 +66,8 @@ args: path: hello.sh start: 0 end: -1 - thoughts: Thoughts: - To begin solving the task, I need to create a new shell script file named 'hello.sh'. This file should contain a script that prints the message 'hello' when executed. I will start by writing the initial content of this script. + thought: Thoughts: + To solve the task of writing a shell script named 'hello.sh' that prints 'hello', I need to create a new file named 'hello.sh'. I will then write the necessary shell script code into this file. Action: @@ -84,8 +84,8 @@ args: path: hello.sh start: 0 end: -1 - thoughts: Thoughts: - Since the file 'hello.sh' has been created, the next step is to write the shell script code that prints "hello". This involves adding a shebang line to specify the shell interpreter and a print command. I will add these lines to the 'hello.sh' file. + thought: Thoughts: + The task is to write a shell script named 'hello.sh' that prints 'hello'. I have already created the file 'hello.sh' in a previous step. Now, I need to write the shell script code into this file. The script should include the shebang line to specify the shell interpreter and a command to print 'hello'. Action: diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_004.log b/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_004.log index 444383d84f..f93d78d595 100644 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_004.log +++ b/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_004.log @@ -66,8 +66,8 @@ args: path: hello.sh start: 0 end: -1 - thoughts: Thoughts: - To begin solving the task, I need to create a new shell script file named 'hello.sh'. This file should contain a script that prints the message 'hello' when executed. I will start by writing the initial content of this script. + thought: Thoughts: + To solve the task of writing a shell script named 'hello.sh' that prints 'hello', I need to create a new file named 'hello.sh'. I will then write the necessary shell script code into this file. Action: @@ -84,8 +84,8 @@ args: path: hello.sh start: 0 end: -1 - thoughts: Thoughts: - Since the file 'hello.sh' has been created, the next step is to write the shell script code that prints "hello". This involves adding a shebang line to specify the shell interpreter and a print command. I will add these lines to the 'hello.sh' file. + thought: Thoughts: + The task is to write a shell script named 'hello.sh' that prints 'hello'. I have already created the file 'hello.sh' in a previous step. Now, I need to write the shell script code into this file. The script should include the shebang line to specify the shell interpreter and a command to print 'hello'. Action: @@ -97,19 +97,21 @@ extras: Memory 2: Previous Action: -action: write +action: read args: path: hello.sh start: 0 end: -1 - thoughts: Thoughts: - I have created the file 'hello.sh' but have not yet written the script content that prints "hello". The next step is to write the shell script code into 'hello.sh'. This will include a shebang line to specify the shell interpreter (bash in this case) and an echo command to print "hello". + thought: Thoughts: + I need to verify the contents of the 'hello.sh' script to ensure that it correctly contains the shell script code to print 'hello'. This involves checking if the shebang line is present and if the print command is correctly written. If the script is correct, I can then proceed to test it to confirm its functionality. Action: Output from Action: -observation: write -content: +observation: read +content: #!/bin/bash +echo "hello" + extras: path: hello.sh ======= End Actions ======= diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_005.log b/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_005.log index a79e0c6a02..f7e8564066 100644 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_005.log +++ b/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_005.log @@ -66,8 +66,8 @@ args: path: hello.sh start: 0 end: -1 - thoughts: Thoughts: - To begin solving the task, I need to create a new shell script file named 'hello.sh'. This file should contain a script that prints the message 'hello' when executed. I will start by writing the initial content of this script. + thought: Thoughts: + To solve the task of writing a shell script named 'hello.sh' that prints 'hello', I need to create a new file named 'hello.sh'. I will then write the necessary shell script code into this file. Action: @@ -84,8 +84,8 @@ args: path: hello.sh start: 0 end: -1 - thoughts: Thoughts: - Since the file 'hello.sh' has been created, the next step is to write the shell script code that prints "hello". This involves adding a shebang line to specify the shell interpreter and a print command. I will add these lines to the 'hello.sh' file. + thought: Thoughts: + The task is to write a shell script named 'hello.sh' that prints 'hello'. I have already created the file 'hello.sh' in a previous step. Now, I need to write the shell script code into this file. The script should include the shebang line to specify the shell interpreter and a command to print 'hello'. Action: @@ -97,31 +97,13 @@ extras: Memory 2: Previous Action: -action: write -args: - path: hello.sh - start: 0 - end: -1 - thoughts: Thoughts: - I have created the file 'hello.sh' but have not yet written the script content that prints "hello". The next step is to write the shell script code into 'hello.sh'. This will include a shebang line to specify the shell interpreter (bash in this case) and an echo command to print "hello". - - Action: - -Output from Action: -observation: write -content: -extras: - path: hello.sh - -Memory 3: -Previous Action: action: read args: path: hello.sh start: 0 end: -1 - thoughts: Thoughts: - I have created the file 'hello.sh' but have not yet confirmed the content of the script. The next step is to read the content of 'hello.sh' to ensure that it contains the correct shell script code that prints "hello". This will help verify that the previous write actions were successful and the file is ready for execution. + thought: Thoughts: + I need to verify the contents of the 'hello.sh' script to ensure that it correctly contains the shell script code to print 'hello'. This involves checking if the shebang line is present and if the print command is correctly written. If the script is correct, I can then proceed to test it to confirm its functionality. Action: @@ -132,6 +114,22 @@ echo "hello" extras: path: hello.sh + +Memory 3: +Previous Action: +action: run +args: + command: bash hello.sh + background: False + thought: + +Output from Action: +observation: run +content: hello +extras: + command_id: -1 + command: bash hello.sh + exit_code: 0 ======= End Actions ======= Use these memories to provide additional context to the problem you are solving. Remember that you have already completed these steps so you do not need to perform them again. diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_006.log b/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_006.log deleted file mode 100644 index 4a68e46a11..0000000000 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/prompt_006.log +++ /dev/null @@ -1,198 +0,0 @@ - - ----------- - -SYSTEM INFO: -You am an autonomous coding agent, here to provide solutions for coding issues. -You have been designed to assist you with a wide range of programming tasks, from code editing and debugging to testing and deployment. -You have access to a variety of tools and commands that you can use to help you solve problems efficiently. - -INSTRUCTIONS: -Now, you're going to solve this issue on your own. You can use any bash commands or custom commands you wish to complete your task. Edit all the files you need to and run any checks or tests that you want. -Remember, YOU CAN ONLY ENTER ONE COMMAND AT A TIME. You should always wait for feedback after every command. -When you're satisfied with all of the changes you've made, you can indicate that you are done by running the exit command. -Note however that you cannot use any interactive session commands (e.g. python, vim, node) in this environment, but you can write scripts and run them. E.g. you can write a python script and then run it with `python .py`. - -NOTE ABOUT THE write COMMAND: Indentation really matters! When editing a file, make sure to insert appropriate indentation before each line! - -IMPORTANT TIPS: -1. Reproduce the bug: Always start by trying to replicate the bug that the issue discusses. If the issue includes code for reproducing the bug, we recommend that you re-implement that in your environment and run it to ensure you can reproduce the bug. Then, start trying to fix it. When you think you've fixed the bug, re-run the bug reproduction script to make sure that the issue has indeed been resolved. - If the bug reproduction script does not print anything when it successfully runs, we recommend adding a print("Script completed successfully, no errors.") command at the end of the file, so that you can be sure the script ran fine all the way through. -2. Try different commands: If you run a command and it doesn't work, try running a different command. A command that did not work once will not work the second time unless you modify it. -3. Navigate large files: If you open a file and need to get to an area around a specific line that is not in the first 100 lines, say line 583, you would use the 'read' command like this: 'read 583'. This is a much faster way to read through the file. -4. Handle input files: If the bug reproduction script requires inputting/reading a specific file, such as 'buggy-input.png', and you'd like to understand how to input that file, conduct a search in the existing repository code to see whether someone else has already done that. Do this by running the command: 'search_dir "buggy-input.png"'. If that doesn't work, use the Linux 'find' command. -5. Understand your context: Always make sure to look at the currently open file and the current working directory. The currently open file might be in a different directory than the working directory. -6. Verify your edits: When editing files, it is easy to accidentally specify a wrong line number or to write code with incorrect indentation. Always check the code after you issue an edit to make sure that it reflects what you wanted to accomplish. If it didn't, issue another command to fix it. -7. Thoroughly test your solution: After making any changes to fix a bug, be sure to thoroughly test your solution to ensure the bug has been resolved. Re-run the bug reproduction script and verify that the issue has been addressed. - - -DOCUMENTATION: -It is recommend that you use the commands provided for interacting with files and your directory because they have been specially built for you. -They will make it much easier for you to look at files and make changes. Using these commands will help you be better at your task. -You can open an file by using either the read or write operations. -- If a file already exists you should read it before making any changes. Use the `edit` command to make changes once you have read it. -- If you are creating a new file use the write command. Use the `edit` command to make changes once you have created the new file. - -Commands: -exit - Executed when task is complete -read [] [] - Shows a given file's contents starting from up to . Default: start_line = 0, end_line = -1. By default the whole file will be read. -write [] [] - Modifies a by replacing the current lines between and with . Default start_line = 0 and end_line = -1. Calling this with no line args will replace the whole file. -browse - Returns the text version of any url, this can be useful to look up documentation or finding issues on github -scroll_up - Takes no arguments. This will scroll up and show you the 100 lines above your current lines -scroll_down - Takes no arguments. This will scroll down and show you the 100 lines below your current lines -edit - This will modify lines in the currently open file. use start_line and end_line to designate which lines to change and then write the multiline changes -goto - This will take you directly to a line and show you the 100 lines below it. - - You can use any bash command you need (cd, ls, rm, grep, dir, mv, wget, git, zip, etc.) with their arguments included -pip install - You can use this to import python packages. Make sure you include the correct package name when using this command. -ls - Use the ls command to view all the files in your current directory, this is a good starting point. -NOT ALLOWED - You cannot use interactive commands like python or node -Custom bash commands: -None - - -The following commands require an open file to be used: edit, scroll_up, scroll_down, goto -To modify the current file use 'edit'. To move through the current file use 'goto' or 'scroll_up'/'scroll_down' -when using write and edit do not surround the code with any "" just write the code. - ----------- - -These are your past 4 actions: -===== Previous Actions ===== - -Memory 0: -Previous Action: -action: write -args: - path: hello.sh - start: 0 - end: -1 - thoughts: Thoughts: - Since the file 'hello.sh' has been created, the next step is to write the shell script code that prints "hello". This involves adding a shebang line to specify the shell interpreter and a print command. I will add these lines to the 'hello.sh' file. - - Action: - -Output from Action: -observation: write -content: -extras: - path: hello.sh - -Memory 1: -Previous Action: -action: write -args: - path: hello.sh - start: 0 - end: -1 - thoughts: Thoughts: - I have created the file 'hello.sh' but have not yet written the script content that prints "hello". The next step is to write the shell script code into 'hello.sh'. This will include a shebang line to specify the shell interpreter (bash in this case) and an echo command to print "hello". - - Action: - -Output from Action: -observation: write -content: -extras: - path: hello.sh - -Memory 2: -Previous Action: -action: read -args: - path: hello.sh - start: 0 - end: -1 - thoughts: Thoughts: - I have created the file 'hello.sh' but have not yet confirmed the content of the script. The next step is to read the content of 'hello.sh' to ensure that it contains the correct shell script code that prints "hello". This will help verify that the previous write actions were successful and the file is ready for execution. - - Action: - -Output from Action: -observation: read -content: #!/bin/bash -echo "hello" - -extras: - path: hello.sh - -Memory 3: -Previous Action: -action: run -args: - command: bash hello.sh - background: False - -Output from Action: -observation: run -content: hello -extras: - command_id: -1 - command: bash hello.sh - exit_code: 0 -======= End Actions ======= -Use these memories to provide additional context to the problem you are solving. -Remember that you have already completed these steps so you do not need to perform them again. - ----------- - -RESPONSE FORMAT: -This is the format of the response you will make in order to solve the current issue. -You will be given multiple iterations to complete this task so break it into steps and solve them one by one. - -Your output must contain the following: -- First, thoughts about what your next action should be and plan it out. - - You will have a memory of your thoughts so you can use this to remember things for the next step. - - Use your thoughts to think about what you are currently doing, what you have done on prior steps and how that relates to solving the problem. -- Second, create a piece of code that will execute your next action based on the thoughts you have. - - Remember that you can only have one action for each thought, do not include multiple actions. - -Your code MUST be surrounded in triple back ticks EXACTLY like this: -``` - -``` - -Notes: -- Adhere to the format so that the program loop continues smoothly, it is very important to only give one command per output. -- DO NOT give more than one command within the triple backticks. This will just throw an error and nothing will happen as a result. -- Do not give multiple code blocks, if you do only the second one will be captured and run, this might give an error if the first one was necessary. -- To execute multiple commands you should write them down in your thoughts section so you can remember it on the next step and execute them then. -- The only commands you are not capable of executing are interactive commands like `python` or `node` by themselves. -- If you think that you have completed the task that has been given to you based on your previous actions and outputs then use ``` exit ``` as the command to let the system know that you are done. -- DO NOT make any copies of your previous memories those will be provided to you at each step, making copies just wastes time and energy. Think smarter not harder. -- The write and edit commands requires proper indentation in the content section ex. `write hw.py def hello(): - print('Hello World')` this is how you would have to format your write command. - - The white spaces matter as the code changes will be added to the code so they must have proper syntax. - -This is a template using the format described above -Items in <> are suggestions for you, fill them out based on the context of the problem you are solving. - -[ FORMAT ] -Thoughts: - - - -Action: -``` - -``` -[ END FORMAT ] - -Do not provide anything extra just your thought and action. - -You are currently trying to complete this task: -Write a shell script 'hello.sh' that prints 'hello'. - -CURRENT WORKSPACE: - Open File: hello.sh on line 0 - You can use these commands with the current file: - Navigation: `scroll_up`, `scroll_down`, and `goto ` - Modification: `edit ` - - -Keep all of the guidelines above in mind when you are thinking and making code. -Please come up with a thought and action based on your current task and latest steps. -Make sure that you do not repeat the same actions, there will not be any changes in result if you do not changes anything. -Be very strict about the formatting that you use and make sure you follow the guidelines. -NEVER output multiple commands. ONLY take ONE STEP at a time. -When you have completed your task run the "exit" command. -Begin with your thought about the next step and then come up with an action to perform your thought. diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/response_001.log b/tests/integration/mock/SWEAgent/test_write_simple_script/response_001.log index 757ced6842..c434807f55 100644 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/response_001.log +++ b/tests/integration/mock/SWEAgent/test_write_simple_script/response_001.log @@ -1,8 +1,8 @@ Thoughts: -To begin solving the task, I need to create a new shell script file named 'hello.sh'. This file should contain a script that prints the message 'hello' when executed. I will start by writing the initial content of this script. +To solve the task of writing a shell script named 'hello.sh' that prints 'hello', I need to create a new file named 'hello.sh'. I will then write the necessary shell script code into this file. Action: ``` write hello.sh #!/bin/bash -echo 'hello' +echo "hello" ``` diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/response_002.log b/tests/integration/mock/SWEAgent/test_write_simple_script/response_002.log index a9b34f581c..1cc79f854b 100644 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/response_002.log +++ b/tests/integration/mock/SWEAgent/test_write_simple_script/response_002.log @@ -1,5 +1,5 @@ Thoughts: -Since the file 'hello.sh' has been created, the next step is to write the shell script code that prints "hello". This involves adding a shebang line to specify the shell interpreter and a print command. I will add these lines to the 'hello.sh' file. +The task is to write a shell script named 'hello.sh' that prints 'hello'. I have already created the file 'hello.sh' in a previous step. Now, I need to write the shell script code into this file. The script should include the shebang line to specify the shell interpreter and a command to print 'hello'. Action: ``` diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/response_003.log b/tests/integration/mock/SWEAgent/test_write_simple_script/response_003.log index 0fa2854495..49592827e5 100644 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/response_003.log +++ b/tests/integration/mock/SWEAgent/test_write_simple_script/response_003.log @@ -1,8 +1,7 @@ Thoughts: -I have created the file 'hello.sh' but have not yet written the script content that prints "hello". The next step is to write the shell script code into 'hello.sh'. This will include a shebang line to specify the shell interpreter (bash in this case) and an echo command to print "hello". +I need to verify the contents of the 'hello.sh' script to ensure that it correctly contains the shell script code to print 'hello'. This involves checking if the shebang line is present and if the print command is correctly written. If the script is correct, I can then proceed to test it to confirm its functionality. Action: ``` -write hello.sh #!/bin/bash -echo "hello" +read hello.sh ``` diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/response_004.log b/tests/integration/mock/SWEAgent/test_write_simple_script/response_004.log index fa6d159a3a..2092717d50 100644 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/response_004.log +++ b/tests/integration/mock/SWEAgent/test_write_simple_script/response_004.log @@ -1,7 +1,7 @@ Thoughts: -I have created the file 'hello.sh' but have not yet confirmed the content of the script. The next step is to read the content of 'hello.sh' to ensure that it contains the correct shell script code that prints "hello". This will help verify that the previous write actions were successful and the file is ready for execution. +The shell script 'hello.sh' has been written and contains the necessary commands to print 'hello'. The next step is to test the script to ensure it functions as expected. This involves running the script and checking the output. If the script runs successfully and prints 'hello', then the task is completed. Action: ``` -read hello.sh +bash hello.sh ``` diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/response_005.log b/tests/integration/mock/SWEAgent/test_write_simple_script/response_005.log index 5ff2634c95..7515501080 100644 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/response_005.log +++ b/tests/integration/mock/SWEAgent/test_write_simple_script/response_005.log @@ -1,7 +1,7 @@ Thoughts: -The shell script 'hello.sh' has been successfully created and contains the necessary code to print "hello". The next logical step is to test the script to ensure it functions as expected. This involves executing the script and observing the output. If the script runs correctly and prints "hello", then the task is completed successfully. +The shell script 'hello.sh' has been successfully created and tested, and it correctly prints 'hello' as required. The script includes the shebang line and the echo command to print 'hello'. Since the script is functioning as expected and meets the task requirements, there are no further modifications or tests needed. It's time to conclude this task. Action: ``` -bash hello.sh +exit ``` diff --git a/tests/integration/mock/SWEAgent/test_write_simple_script/response_006.log b/tests/integration/mock/SWEAgent/test_write_simple_script/response_006.log deleted file mode 100644 index bc1cfb2181..0000000000 --- a/tests/integration/mock/SWEAgent/test_write_simple_script/response_006.log +++ /dev/null @@ -1,7 +0,0 @@ -Thoughts: -The shell script 'hello.sh' has been successfully created and tested, and it correctly prints "hello" when executed. Since the task was to write a shell script that prints 'hello' and this has been achieved as evidenced by the output from the previous run command, there are no further modifications or actions needed on the script. Therefore, I should conclude this task. - -Action: -``` -exit -``` diff --git a/tests/integration/test_agent.py b/tests/integration/test_agent.py index d1f55e9eaa..2a3522d658 100644 --- a/tests/integration/test_agent.py +++ b/tests/integration/test_agent.py @@ -7,7 +7,11 @@ import pytest from opendevin.main import main -@pytest.mark.skipif(os.environ.get('AGENT') == 'CodeActAgent', reason='CodeActAgent requires task to be in a special format') +# skip if +@pytest.mark.skipif( + os.getenv('AGENT') == 'CodeActAgent' and os.getenv('SANDBOX_TYPE').lower() == 'exec', + reason='CodeActAgent does not support exec sandbox since exec sandbox is NOT stateful' +) def test_write_simple_script(): task = "Write a shell script 'hello.sh' that prints 'hello'." asyncio.run(main(task)) diff --git a/tests/unit/test_action_serialization.py b/tests/unit/test_action_serialization.py index 4f7f733ca5..33632e2820 100644 --- a/tests/unit/test_action_serialization.py +++ b/tests/unit/test_action_serialization.py @@ -39,7 +39,7 @@ def test_agent_think_action_serialization_deserialization(): def test_agent_recall_action_serialization_deserialization(): original_action_dict = { 'action': 'recall', - 'args': {'query': 'Test query.'} + 'args': {'query': 'Test query.', 'thought': ''} } serialization_deserialization(original_action_dict, AgentRecallAction) @@ -47,7 +47,7 @@ def test_agent_recall_action_serialization_deserialization(): def test_agent_finish_action_serialization_deserialization(): original_action_dict = { 'action': 'finish', - 'args': {'outputs': {}}, + 'args': {'outputs': {}, 'thought': ''} } serialization_deserialization(original_action_dict, AgentFinishAction) @@ -55,7 +55,7 @@ def test_agent_finish_action_serialization_deserialization(): def test_cmd_kill_action_serialization_deserialization(): original_action_dict = { 'action': 'kill', - 'args': {'id': '1337'} + 'args': {'id': '1337', 'thought': ''} } serialization_deserialization(original_action_dict, CmdKillAction) @@ -63,7 +63,7 @@ def test_cmd_kill_action_serialization_deserialization(): def test_cmd_run_action_serialization_deserialization(): original_action_dict = { 'action': 'run', - 'args': {'command': 'echo "Hello world"', 'background': True} + 'args': {'command': 'echo "Hello world"', 'background': True, 'thought': ''} } serialization_deserialization(original_action_dict, CmdRunAction) @@ -71,7 +71,7 @@ def test_cmd_run_action_serialization_deserialization(): def test_browse_url_action_serialization_deserialization(): original_action_dict = { 'action': 'browse', - 'args': {'url': 'https://www.example.com'} + 'args': {'thought': '', 'url': 'https://www.example.com'} } serialization_deserialization(original_action_dict, BrowseURLAction) @@ -87,7 +87,7 @@ def test_github_push_action_serialization_deserialization(): def test_file_read_action_serialization_deserialization(): original_action_dict = { 'action': 'read', - 'args': {'path': '/path/to/file.txt', 'start': 0, 'end': -1, 'thoughts': 'None'} + 'args': {'path': '/path/to/file.txt', 'start': 0, 'end': -1, 'thought': 'None'} } serialization_deserialization(original_action_dict, FileReadAction) @@ -95,7 +95,7 @@ def test_file_read_action_serialization_deserialization(): def test_file_write_action_serialization_deserialization(): original_action_dict = { 'action': 'write', - 'args': {'path': '/path/to/file.txt', 'content': 'Hello world', 'start': 0, 'end': 1, 'thoughts': 'None'} + 'args': {'path': '/path/to/file.txt', 'content': 'Hello world', 'start': 0, 'end': 1, 'thought': 'None'} } serialization_deserialization(original_action_dict, FileWriteAction) @@ -103,7 +103,7 @@ def test_file_write_action_serialization_deserialization(): def test_add_task_action_serialization_deserialization(): original_action_dict = { 'action': 'add_task', - 'args': {'parent': 'Test parent', 'goal': 'Test goal', 'subtasks': []} + 'args': {'parent': 'Test parent', 'goal': 'Test goal', 'subtasks': [], 'thought': ''} } serialization_deserialization(original_action_dict, AddTaskAction) @@ -111,6 +111,6 @@ def test_add_task_action_serialization_deserialization(): def test_modify_task_action_serialization_deserialization(): original_action_dict = { 'action': 'modify_task', - 'args': {'id': 1, 'state': 'Test state.'} + 'args': {'id': 1, 'state': 'Test state.', 'thought': ''} } serialization_deserialization(original_action_dict, ModifyTaskAction)