mini-swe-agent 1.1.1__py3-none-any.whl → 1.2.0__py3-none-any.whl

{mini_swe_agent-1.1.1.dist-info → mini_swe_agent-1.2.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mini-swe-agent
-Version: 1.1.1
+Version: 1.2.0
 Summary: Nano SWE Agent - A simple AI software engineering agent
 Author-email: Kilian Lieret <kilian.lieret@posteo.de>, "Carlos E. Jimenez" <carlosej@princeton.edu>
 License: MIT License
@@ -32,9 +32,9 @@ Classifier: Development Status :: 3 - Alpha
 Classifier: Operating System :: OS Independent
 Classifier: Intended Audience :: Developers
 Classifier: License :: OSI Approved :: MIT License
-Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3 :: Only
-Requires-Python: >=3.11
+Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE.md
 Requires-Dist: pyyaml
@@ -67,8 +67,8 @@ Requires-Dist: mkdocs-glightbox; extra == "dev"
 Dynamic: license-file
 
 <div align="center">
-
 <a href="https://mini-swe-agent.com/latest/"><img src="https://github.com/SWE-agent/mini-swe-agent/raw/main/docs/assets/mini-swe-agent-banner.svg" alt="mini-swe-agent banner" style="height: 7em"/></a>
+</div>
 
 <h1>The 100 line AI agent that solves GitHub issues & more</h1>
 
@@ -76,27 +76,25 @@ Dynamic: license-file
 [![Slack](https://img.shields.io/badge/Slack-4A154B?style=for-the-badge&logo=slack&logoColor=white)](https://join.slack.com/t/swe-bench/shared_invite/zt-36pj9bu5s-o3_yXPZbaH2wVnxnss1EkQ)
 [![PyPI - Version](https://img.shields.io/pypi/v/mini-swe-agent?style=for-the-badge&logo=python&logoColor=white&labelColor=black&color=deeppink)](https://pypi.org/project/mini-swe-agent/)
 
-</div>
-
-In 2024, [SWE-bench](https://github.com/swe-bench/SWE-bench) & [SWE-agent](https://github.com/swe-agent/swe-agent) helped kickstart the agentic AI for software revolution.
+In 2024, [SWE-bench](https://github.com/swe-bench/SWE-bench) & [SWE-agent](https://github.com/swe-agent/swe-agent) helped kickstart the coding agent revolution.
 
 We now ask: **What if SWE-agent was 100x smaller, and still worked nearly as well?**
 
 `mini` is for
 
-- 🧪 **Researchers** who want to **benchmark, fine-tune or RL** without assumptions, bloat, or surprises
-- 🧑‍💻 **Hackers & power users** who like their tools like their scripts: **short, sharp, and readable**
-- 🐳 **Engineers** who want something **trivial to sandbox & to deploy anywhere**
+- **Researchers** who want to **benchmark, fine-tune or RL** without assumptions, bloat, or surprises
+- **Developers** who like their tools like their scripts: **short, sharp, and readable**
+- **Engineers** who want something **trivial to sandbox & to deploy anywhere**
 
 Here's some details:
 
-- **🐜 Minimal**: Just [100 lines of python](https://github.com/SWE-agent/mini-swe-agent/blob/main/src/minisweagent/agents/default.py) (+100 total for [env](https://github.com/SWE-agent/mini-swe-agent/blob/main/src/minisweagent/environments/local.py),
+- **Minimal**: Just [100 lines of python](https://github.com/SWE-agent/mini-swe-agent/blob/main/src/minisweagent/agents/default.py) (+100 total for [env](https://github.com/SWE-agent/mini-swe-agent/blob/main/src/minisweagent/environments/local.py),
 [model](https://github.com/SWE-agent/mini-swe-agent/blob/main/src/minisweagent/models/litellm_model.py), [script](https://github.com/SWE-agent/mini-swe-agent/blob/main/src/minisweagent/run/hello_world.py)) — no fancy dependencies!
-- **💪 Powerful:** Resolves 65% of GitHub issues in the [SWE-bench verified benchmark](https://www.swebench.com/).
-- **🤗 Friendly:** Comes with **two convenient UIs** that will turn this into your daily dev swiss army knife!
-- **🍀 Environments:** In addition to local envs, you can use **docker**, **podman**, **singularity**, **apptainer**, and more
-- **🧪 Tested:** [![Codecov](https://img.shields.io/codecov/c/github/swe-agent/mini-swe-agent?style=flat-square)](https://codecov.io/gh/SWE-agent/mini-swe-agent)
-- **🎓 Cutting edge:** Built by the Princeton & Stanford team behind [SWE-bench](https://swebench.com) and [SWE-agent](https://swe-agent.com).
+- **Powerful:** Resolves 65% of GitHub issues in the [SWE-bench verified benchmark](https://www.swebench.com/) (with Claude Sonnet 4).
+- **Convenient:** Comes with UIs that turn this into your daily dev swiss army knife!
+- **Deployable:** In addition to local envs, you can use **docker**, **podman**, **singularity**, **apptainer**, and more
+- **Tested:** [![Codecov](https://img.shields.io/codecov/c/github/swe-agent/mini-swe-agent?style=flat-square)](https://codecov.io/gh/SWE-agent/mini-swe-agent)
+- **Cutting edge:** Built by the Princeton & Stanford team behind [SWE-bench](https://swebench.com) and [SWE-agent](https://swe-agent.com).
 
 <details>
 
@@ -106,14 +104,15 @@ Here's some details:
 However, one year later, as LMs have become more capable, a lot of this is not needed at all to build a useful agent!
 In fact, mini-SWE-agent
 
-- Does not have any tools other than bash — it doesn't even use the tool-calling interface of the LMs.
-  This means that you can run it with literally any model. When running in sandboxed environments you also don't need to to take care
+- **Does not have any tools other than bash** — it doesn't even use the tool-calling interface of the LMs.
+  This means that you can run it with literally any model. When running in sandboxed environments you also don't need to take care
   of installing a single package — all it needs is bash.
-- Has a completely linear history — every step of the agent just appends to the messages and that's it.
+- **Has a completely linear history** — every step of the agent just appends to the messages and that's it.
   So there's no difference between the trajectory and the messages that you pass on to the LM.
-- Executes actions with `subprocess.run` — every action is completely independent (as opposed to keeping a stateful shell session running).
+  Great for debugging & fine-tuning.
+- **Executes actions with `subprocess.run`** — every action is completely independent (as opposed to keeping a stateful shell session running).
   This makes it trivial to execute the actions in sandboxes (literally just switch out `subprocess.run` with `docker exec`) and to
-  scale up effortlessly.
+  scale up effortlessly. Seriously, this is [a big deal](https://mini-swe-agent.com/latest/faq/#why-no-shell-session), trust me.
 
 This makes it perfect as a baseline system and for a system that puts the language model (rather than
 the agent scaffold) in the middle of our attention.
@@ -123,41 +122,42 @@ the agent scaffold) in the middle of our attention.
 <details>
 <summary>More motivation (as a tool)</summary>
 
-Some agents are overfitted research artifacts.
-Others are UI-heavy tools, highly optimized for a specific user experience.
-Both variants are hard to understand.
+Some agents are overfitted research artifacts. Others are UI-heavy frontend monsters.
 
-`mini` strives to be
+`mini` wants to be a hackable tool, not a black box.
 
 - **Simple** enough to understand at a glance
 - **Convenient** enough to use in daily workflows
 - **Flexible** to extend
 
-A hackable tool, not a black box.
-
-Unlike other agents (including our own [swe-agent](https://swe-agent.com/latest/)),
-it is radically simpler, because it
+Unlike other agents (including our own [swe-agent](https://swe-agent.com/latest/)), it is radically simpler, because it:
 
-- Does not have any tools other than bash — it doesn't even use the tool-calling interface of the LMs.
-- Has a completely linear history — every step of the agent just appends to the messages and that's it.
-- Executes actions with `subprocess.run` — every action is completely independent (as opposed to keeping a stateful shell session running).
+- **Does not have any tools other than bash** — it doesn't even use the tool-calling interface of the LMs.
+  Instead of implementing custom tools for every specific thing the agent might want to do, the focus is fully on the LM utilizing the shell to its full potential.
+  Want it to do something specific like opening a PR?
+  Just tell the LM to figure it out rather than spending time to implement it in the agent.
+- **Executes actions with `subprocess.run`** — every action is completely independent (as opposed to keeping a stateful shell session running).
+  This is [a big deal](https://mini-swe-agent.com/latest/faq/#why-no-shell-session) for the stability of the agent, trust me.
+- **Has a completely linear history** — every step of the agent just appends to the messages that are passed to the LM in the next step and that's it.
+  This is great for debugging and understanding what the LM is prompted with.
 
 </details>
 
 <details>
 <summary>Should I use SWE-agent or mini-SWE-agent?</summary>
 
-You should use [`swe-agent`](https://swe-agent.com/latest/) if
-
-- You need specific tools or want to experiment with different tools
-- You want to experiment with different history processors
-- You want very powerful yaml configuration without touching code
-
-You should use [`mini-swe-agent`](https://mini-swe-agent.com/latest/) if
+You should use `mini-swe-agent` if
 
 - You want a quick command line tool that works locally
 - You want an agent with a very simple control flow
 - You want even faster, simpler & more stable sandboxing & benchmark evaluations
+- You are doing FT or RL and don't want to overfit to a specific agent scaffold
+
+You should use `swe-agent` if
+
+- You need specific tools or want to experiment with different tools
+- You want to experiment with different history processors
+- You want very powerful yaml configuration without touching code
 
 What you get with both
 
@@ -240,21 +240,23 @@ agent.run("Write a sudoku game")
 </tr>
 </table>
 
-## 🔥 Let's get started!
+## Let's get started!
 
-Install + run in virtual environment
+Option 1: Install + run in virtual environment
 
 ```bash
+pip install uv && uvx mini-swe-agent [-v]
+# or
 pip install pipx && pipx ensurepath && pipx run mini-swe-agent [-v]
 ```
 
-Alternative: Install in current environment
+Option 2: Install in current environment
 
 ```bash
 pip install mini-swe-agent && mini [-v]
 ```
 
-Alternative: Install from source
+Option 3: Install from source
 
 ```bash
 git clone https://github.com/SWE-agent/mini-swe-agent.git
@@ -272,7 +274,21 @@ Read more in our [documentation](https://mini-swe-agent.com/latest/):
 * [FAQ](https://mini-swe-agent.com/latest/faq/)
 * [Contribute!](https://mini-swe-agent.com/latest/contributing/)
 
-## 👀 More agentic AI
+## Bottom line
+
+If you found this work helpful, please consider citing the [SWE-agent paper](https://arxiv.org/abs/2405.15793) in your work:
+
+```bibtex
+@inproceedings{yang2024sweagent,
+  title={{SWE}-agent: Agent-Computer Interfaces Enable Automated Software Engineering},
+  author={John Yang and Carlos E Jimenez and Alexander Wettig and Kilian Lieret and Shunyu Yao and Karthik R Narasimhan and Ofir Press},
+  booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
+  year={2024},
+  url={https://arxiv.org/abs/2405.15793}
+}
+```
+
+More agentic AI:
 
 <div align="center">
   <a href="https://github.com/SWE-agent/SWE-agent"><img src="https://github.com/SWE-agent/mini-swe-agent/raw/main/docs/assets/sweagent_logo_text_below.svg" alt="SWE-agent" height="120px"></a>

{mini_swe_agent-1.1.1.dist-info → mini_swe_agent-1.2.0.dist-info}/RECORD

@@ -1,17 +1,17 @@
-mini_swe_agent-1.1.1.dist-info/licenses/LICENSE.md,sha256=D3luWPkdHAe7LBsdD4vzqDAXw6Xewb3G-uczss0uh1s,1094
-minisweagent/__init__.py,sha256=Ya6NfAmQ-XUSM7SEr46KdNQIEHt-sfo9TUXD_NvJEzU,1787
+mini_swe_agent-1.2.0.dist-info/licenses/LICENSE.md,sha256=D3luWPkdHAe7LBsdD4vzqDAXw6Xewb3G-uczss0uh1s,1094
+minisweagent/__init__.py,sha256=O834rP05yuUp1YEVIdVhl0wEGk2r43BSKUTGg7iACJU,1787
 minisweagent/__main__.py,sha256=FIyAOiw--c3FQ2g240FOM1FdL0lk_PxSpixu0pQ7WFo,194
 minisweagent/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 minisweagent/agents/__init__.py,sha256=cpjJLzg1IGxLM-tZpoMJV9S33ye13XtdBO0x7DU_Lrk,48
 minisweagent/agents/default.py,sha256=6TZUfKch6e_7m05BXmjqUjV-k068s6ldg7T6zJBbQW8,5438
-minisweagent/agents/interactive.py,sha256=w2lCVUItq7EMCAnPJu0q7RqzU-ScQCz1HNulGGDG8pE,7146
+minisweagent/agents/interactive.py,sha256=_DCBabdwdIR4gAojT_TaQW2MSFtmBq997mwmiDGYdRA,7327
 minisweagent/agents/interactive_textual.py,sha256=zzNsq1OkEmrBFVK3t1dxnrE7W7xU7Vc-WN47bxZcDIk,12657
 minisweagent/config/README.md,sha256=tPruhnQDhZ8ugc1FNPKk9tVMRltmmIjdYgvHCmN-3Hs,354
 minisweagent/config/__init__.py,sha256=UfORdQID1Ek_dduZlybUsIKJjihImkSqNU5tIjpw0hk,694
 minisweagent/config/default.yaml,sha256=AGhcIq6X6n5Fs71ufO3B6CtZ4PS877tCxkPkrWR5Ylg,4497
 minisweagent/config/github_issue.yaml,sha256=evvu3AJ52tXYSdami9_B8zfazOAE2r2XXkzVmScBoKc,4539
 minisweagent/config/mini.tcss,sha256=ThSOtS6JpXxqEYGX69TLy6gPZzuijngsNLI6SjnEJLY,1821
-minisweagent/config/mini.yaml,sha256=zmju4EUcybXlHOSzj-ZL2Se1803nXsZPk6jgrB9KxYw,4815
+minisweagent/config/mini.yaml,sha256=U-mTAgnrT2mn9_VxKSnLlq26Einxq0grqiY3esTtmGg,4983
 minisweagent/config/extra/__init__.py,sha256=e1MoAlDn_wc9HnXNoncf1P-B4DQ-iRf6n7Q_txjZGRI,52
 minisweagent/config/extra/swebench.yaml,sha256=LNpTahpul6HL0HozgAAz-C6kpX3wZA7Tg8uE-ZmgrF4,7577
 minisweagent/environments/__init__.py,sha256=g5mKac1YgVOZVKvmiAiuyPSevRYpI69V4vYrbCH3gsI,54
@@ -30,17 +30,17 @@ minisweagent/run/__init__.py,sha256=WIoYgHVl7iZF2YncrfV3IttupG6P5KogroKHKECka3A,
 minisweagent/run/github_issue.py,sha256=GWOkGM09jOYV93p6xIM_kKWmC1yP_d5lprafWlqoBN0,2748
 minisweagent/run/hello_world.py,sha256=erLnEwNmPFLxq3-8zyv66Vy1kIqMqQf97vISX7LrQXg,959
 minisweagent/run/inspector.py,sha256=QnY3oYzm-yq3w9Jzs112Lco2Rg84vSocAWrQRVz_1lc,7127
-minisweagent/run/mini.py,sha256=8fKSss-FQYJektvysLKwRWLNVBhJFpSjNn95xAj8EZM,4265
+minisweagent/run/mini.py,sha256=9EzTUT1cra6sHkqUTGu5oqx9Esgt_XOSGW9cxFyXLd8,4339
 minisweagent/run/mini_extra.py,sha256=ecA1PnTWElpO60G9RktvVLtUOf3bZ_ESmnSttS6izhQ,1465
 minisweagent/run/extra/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-minisweagent/run/extra/config.py,sha256=Xf-JKpD4S_yEhiqh7AZWPEGYVu5LrBnpf9teiAoz-wI,3117
+minisweagent/run/extra/config.py,sha256=paMHfplhKsqNmzhCmozxhXWHvBzBCUlwUWD8N7ytCPc,3277
 minisweagent/run/extra/swebench.py,sha256=x50nFj4tdXkdoad6TkK7tP8CSgf-WkrY0IdSMe8_oX0,9564
 minisweagent/run/extra/swebench_single.py,sha256=L3Kk4G65o3MCPLMEwGNIs77-AFf6Lfc8o1oxrbN-ZWM,1991
 minisweagent/run/extra/utils/batch_progress.py,sha256=u__khJ-fipZLxTJu43LamGAtPUCqEZYEi8J7SfH7X6A,6211
 minisweagent/run/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 minisweagent/run/utils/save.py,sha256=3_kuutw-uAGIhEoDawA3_FPeSz1vWuCWpJl80j5u7_s,893
-mini_swe_agent-1.1.1.dist-info/METADATA,sha256=cUtVzc_xyhGvM3VZ2QK3oD_horO54bpZQ1XLx5fSdf4,12281
-mini_swe_agent-1.1.1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-mini_swe_agent-1.1.1.dist-info/entry_points.txt,sha256=d1_yRbTaGjs1UXHa6JQK0sKDGBIVGm8oeW0k2kfbJgQ,182
-mini_swe_agent-1.1.1.dist-info/top_level.txt,sha256=zKF4t8bFpV87fdVABZt2Da-vnb4Vkh_CxkwQx5YT4Ew,13
-mini_swe_agent-1.1.1.dist-info/RECORD,,
+mini_swe_agent-1.2.0.dist-info/METADATA,sha256=4JI2AGHnBGJhAtkSFhmTBwH7-2QO0ymepR2g8TQGaAA,13459
+mini_swe_agent-1.2.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+mini_swe_agent-1.2.0.dist-info/entry_points.txt,sha256=d1_yRbTaGjs1UXHa6JQK0sKDGBIVGm8oeW0k2kfbJgQ,182
+mini_swe_agent-1.2.0.dist-info/top_level.txt,sha256=zKF4t8bFpV87fdVABZt2Da-vnb4Vkh_CxkwQx5YT4Ew,13
+mini_swe_agent-1.2.0.dist-info/RECORD,,

minisweagent/__init__.py

@@ -8,7 +8,7 @@ This file provides:
   unless you want the static type checking.
 """
 
-__version__ = "1.1.1"
+__version__ = "1.2.0"
 
 import os
 from pathlib import Path

minisweagent/agents/interactive.py

@@ -28,6 +28,8 @@ class InteractiveAgentConfig(AgentConfig):
     """Whether to confirm actions."""
     whitelist_actions: list[str] = field(default_factory=list)
     """Never confirm actions that match these regular expressions."""
+    confirm_exit: bool = True
+    """If the agent wants to finish, do we ask for confirmation from user?"""
 
 
 class InteractiveAgent(DefaultAgent):
@@ -137,12 +139,13 @@ class InteractiveAgent(DefaultAgent):
         try:
             return super().has_finished(output)
         except Submitted as e:
-            console.print(
-                "[bold green]Agent wants to finish.[/bold green] "
-                "[green]Type a comment to give it a new task or press enter to quit.\n"
-                "[bold yellow]>[/bold yellow] ",
-                end="",
-            )
-            if new_task := self._prompt_and_handle_special("").strip():
-                raise NonTerminatingException(f"The user added a new task: {new_task}")
+            if self.config.confirm_exit:
+                console.print(
+                    "[bold green]Agent wants to finish.[/bold green] "
+                    "[green]Type a comment to give it a new task or press enter to quit.\n"
+                    "[bold yellow]>[/bold yellow] ",
+                    end="",
+                )
+                if new_task := self._prompt_and_handle_special("").strip():
+                    raise NonTerminatingException(f"The user added a new task: {new_task}")
             raise e

minisweagent/config/mini.yaml

@@ -23,7 +23,9 @@ agent:
     You can execute bash commands and edit files to implement the necessary changes.
 
     ## Recommended Workflow
-    1. Analyze the codebase by finding and reading relevant files
+    1. Analyze the codebase by finding and reading relevant files.
+       If present, you might want to take a look at the following files that set additional guidelines
+       for your work: CLAUDE.md, .cursor/rules/<relevant rules>
     2. Create a script to reproduce the issue
     3. Edit the source code to resolve the issue
     4. Verify your fix works by running your script again

minisweagent/run/extra/config.py

@@ -58,6 +58,9 @@ def setup():
     ).strip()
     if default_model:
         set_key(global_config_file, "MSWEA_MODEL_NAME", default_model)
+    console.print(
+        "[bold yellow]If you already have your API keys set as environment variables, you can ignore the next question.[/bold yellow]"
+    )
     key_name = prompt("Enter your API key name (e.g., ANTHROPIC_API_KEY): ").strip()
     key_value = None
     if key_name:

minisweagent/run/mini.py

@@ -48,8 +48,6 @@ def run_interactive(model: Model, env: Environment, agent_config: dict, task: st
     exit_status, result = None, None
     try:
         exit_status, result = agent.run(task)
-    except KeyboardInterrupt:
-        console.print("\n[bold red]KeyboardInterrupt -- goodbye[/bold red]")
     finally:
         if output:
             save_traj(agent, output, exit_status=exit_status, result=result)
@@ -65,10 +63,9 @@ def run_textual(model: Model, env: Environment, agent_config: dict, task: str, o
     )
     try:
         agent_app.run()
-    except KeyboardInterrupt:
-        typer.echo("\nKeyboardInterrupt -- goodbye")
     finally:
-        save_traj(agent_app.agent, Path("traj.json"), exit_status=agent_app.exit_status, result=agent_app.result)
+        if output:
+            save_traj(agent_app.agent, output, exit_status=agent_app.exit_status, result=agent_app.result)
 
 
 @app.command(help=__doc__)
@@ -85,6 +82,9 @@ def main(
     cost_limit: float | None = typer.Option(None, "-l", "--cost-limit", help="Cost limit. Set to 0 to disable."),
     config_spec: Path = typer.Option(DEFAULT_CONFIG, "-c", "--config", help="Path to config file"),
     output: Path | None = typer.Option(None, "-o", "--output", help="Output file"),
+    exit_immediately: bool = typer.Option(
+        False, "--exit-immediately", help="Exit immediately when the agent wants to finish instead of prompting."
+    ),
 ) -> Any:
     configure_if_first_time()
     config = yaml.safe_load(get_config_path(config_spec).read_text())
@@ -105,6 +105,8 @@ def main(
     config["agent"]["mode"] = "confirm" if not yolo else "yolo"
     if cost_limit:
         config["agent"]["cost_limit"] = cost_limit
+    if not visual and exit_immediately:
+        config["agent"]["confirm_exit"] = False
     model = get_model(model_name, config.get("model", {}))
     env = LocalEnvironment(**config.get("env", {}))