Quickstart Fixes (#1)

* Changed to correct import

* Added null checks to rlm core

* Added example to quickstart script

* Move quickstart to examples folder

* Remove uv.lock changes

* Merged rlm_example and quickstart

* Wrap model reference in _to_str

* Update README with quickstart example

* Refine Quick Setup instructions in README

Updated setup instructions for clarity and formatting.

---------

Co-authored-by: az <alex.lx.zhang@gmail.com>

Author: kmad

README.md

@@ -47,12 +47,22 @@ pip install git+https://github.com/alexzhang13/rlm.git
 
 ## Quick Setup
 Set up the dependencies with `uv` (or your virtual environment of choice):
-```
+```bash
 curl -LsSf https://astral.sh/uv/install.sh | sh
 uv init && uv venv --python 3.12  # change version as needed
 uv pip install -e .
 ```
 
+To run a quick test, make sure your `OPENAI_API_KEY` is set properly and run the following:
+```bash
+cd examples/
+uv run quickstart.py
+```
+
+This will generate console output as well as a log which you can use with the visualizer to explore the trajectories. 
+
+
+## REPL Environments
 The default RLM object uses a REPL environment that runs on the host process through Python `exec` calls. It uses the same virtual environment as the host process (i.e. it will have access to the same dependencies), but with some limitations in its available global modules. As an example, we can call RLM completions using GPT-5-nano with the Portkey client:
 ```python
 from rlm import RLM
@@ -66,7 +76,6 @@ rlm = RLM(
 print(rlm.completion("Print me the first 100 powers of two, each on a newline.").response)
 ```
 
-## REPL Environments
 We support two types of REPL environments -- isolated, and non-isolated. Non-isolated environments (default) run code execution on the same machine as the RLM (e.g. through `exec`), which is pretty reasonable for some local low-risk tasks, like simple benchmarking, but can be problematic if the prompts or tool calls can interact with malicious users. Fully isolated environments used Cloud-based sandboxes (e.g. Prime Sandboxes, [Modal Sandboxes](https://modal.com/docs/guide/sandboxes)) to run code generated by the RLM, ensuring completely isolation from the host process. Environments can be added, but we natively support the following: `local` (default), `modal`, `prime`.
 
 ```python

examples/quickstart.py

@@ -0,0 +1,27 @@
+import os
+
+from dotenv import load_dotenv
+
+from rlm import RLM
+from rlm.logger import RLMLogger
+
+load_dotenv()
+
+logger = RLMLogger(log_dir="./logs")
+
+rlm = RLM(
+    backend="openai", # or "portkey", etc.
+    backend_kwargs={
+            "model_name": "gpt-5-nano",
+            "api_key": os.getenv("OPENAI_API_KEY"),
+        },
+    environment="local",
+    environment_kwargs={},
+    max_depth=1,
+    logger=logger,
+    verbose=True,  # For printing to console with rich, disabled by default.
+)
+
+result = rlm.completion("Print me the first 100 powers of two, each on a newline.")
+
+print(result)

examples/rlm_example.py

@@ -8,9 +8,9 @@ description = "Recursive Language Models."
 readme = "README.md"
 requires-python = ">=3.11"
 dependencies = [
-    "dotenv>=0.9.9",
     "openai>=2.14.0",
     "portkey-ai>=2.1.0",
+    "python-dotenv>=1.2.1",
     "requests>=2.32.5",
     "rich>=13.0.0",
 ]

pyproject.toml

@@ -71,7 +71,9 @@ def __init__(
         self.backend = backend
         self.backend_kwargs = backend_kwargs
         self.environment_type = environment
-        self.environment_kwargs = environment_kwargs.copy()
+        self.environment_kwargs = (
+            environment_kwargs.copy() if environment_kwargs is not None else {}
+        )
         self.other_backends = other_backends
         self.other_backend_kwargs = other_backend_kwargs
 
@@ -85,13 +87,17 @@ def __init__(
         # Log metadata if logger is provided
         if self.logger or verbose:
             metadata = RLMMetadata(
-                root_model=backend_kwargs.get("model_name", "unknown"),
+                root_model=backend_kwargs.get("model_name", "unknown")
+                if backend_kwargs
+                else "unknown",
                 max_depth=max_depth,
                 max_iterations=max_iterations,
                 backend=backend,
-                backend_kwargs=filter_sensitive_keys(backend_kwargs),
+                backend_kwargs=filter_sensitive_keys(backend_kwargs) if backend_kwargs else {},
                 environment_type=environment,
-                environment_kwargs=filter_sensitive_keys(environment_kwargs),
+                environment_kwargs=filter_sensitive_keys(environment_kwargs)
+                if environment_kwargs
+                else {},
                 other_backends=other_backends,
             )
             if self.logger:
@@ -196,7 +202,9 @@ def completion(
                     self.verbose.print_final_answer(final_answer)
                     self.verbose.print_summary(i + 1, time_end - time_start, usage.to_dict())
                     return RLMChatCompletion(
-                        root_model=self.backend_kwargs.get("model_name", "unknown"),
+                        root_model=self.backend_kwargs.get("model_name", "unknown")
+                        if self.backend_kwargs
+                        else "unknown",
                         prompt=prompt,
                         response=final_answer,
                         usage_summary=usage,
@@ -216,7 +224,9 @@ def completion(
             self.verbose.print_final_answer(final_answer)
             self.verbose.print_summary(self.max_iterations, time_end - time_start, usage.to_dict())
             return RLMChatCompletion(
-                root_model=self.backend_kwargs.get("model_name", "unknown"),
+                root_model=self.backend_kwargs.get("model_name", "unknown")
+                if self.backend_kwargs
+                else "unknown",
                 prompt=prompt,
                 response=final_answer,
                 usage_summary=usage,

rlm/core/rlm.py

@@ -277,7 +277,7 @@ def print_subcall(
         header = Text()
         header.append("  ↳ ", style=STYLE_SECONDARY)
         header.append("Sub-call: ", style=STYLE_SECONDARY)
-        header.append(model, style=STYLE_ACCENT)
+        header.append(_to_str(model), style=STYLE_ACCENT)
         if execution_time:
             header.append(f"  ({execution_time:.2f}s)", style=STYLE_MUTED)