doc(readme): typos and missing key in .env.template (#4)

* improve readme and scripts.

* update e2b part.

* update section on e2b.

* remove local e2b

* update the code and script.

* remove e2b auth login.

Author: gali-leilei

README.md

@@ -178,25 +178,49 @@ If you wish to use a different LLM as the primary agent model, you will need to
 
 ```bash
 ## copy environment variable template and prepare yours in .env file
+cd miroflow/apps/prepare-benchmark
+cp .env.template .env
+vim .env
+cd miroflow/apps/run-agent
 cp .env.template .env
 vim .env
 ```
 
-**Step 3:** Prepare E2B Sandbox:
+**Step 3:** Prepare E2B Sandbox (Optional)
 
 > [!TIP]
+> We provide a public E2B sandbox template. Follow this step if you want to reproduce.
+>
 > For the E2B sandbox service, we recommend setting up a Linux Docker image with a comprehensive set of apt and Python packages pre-installed. Without these pre-installed packages, the agent will need to spend extra steps and context installing them, resulting in reduced token efficiency.
+>
+> you need to have `npm` install and `docker` running locally.
+
+
+1. Install `e2b` command line and login:
 
-1. Download our pre-configured Dockerfile:
+```shell
+## install e2b
+npm install -g @e2b/cli
+## check that it is available
+which e2b 
+```
+
+2. Download our pre-configured Dockerfile:
 [e2b.Dockerfile](https://github.com/MiroMindAI/MiroFlow/blob/main/docs/e2b.Dockerfile).
 
-2. Run `e2b template build` command [check official doc here](https://e2b.dev/docs/sdk-reference/cli/v1.0.2/template), use `all_pip_apt_pkg` as the name of template.
+```shell
+wget https://github.com/MiroMindAI/MiroFlow/blob/main/docs/e2b.Dockerfile
+```
 
-   Command example:
-   ```shell
-   E2B_ACCESS_TOKEN=${your-token}
-   e2b template build -c "/root/.jupyter/start-up.sh" -t <team_id> -n all_pip_apt_pkg` -d ./e2b.Dockerfile
-   ```
+3. Run `e2b template build` command [check official doc here](https://e2b.dev/docs/sdk-reference/cli/v1.0.2/template), use `all_pip_apt_pkg` as the name of template.
+
+```shell
+## build the template with `docker build` locally
+E2B_ACCESS_TOKEN=${your-token}
+e2b template build -c "/root/.jupyter/start-up.sh" -n "all_pip_apt_pkg" -d ./e2b.Dockerfile
+## check that template is built successfully
+E2B_ACCESS_TOKEN=${your-token} e2b template list
+```
 
 For additional information, please see the [E2B Docker documentation](https://e2b.dev/docs/sandbox-template).
 
@@ -218,7 +242,7 @@ Run prebuilt agent on the benchmark data:
 ```bash
 ## download data
 cd miroflow/apps/prepare-benchmark
-uv run python main.py get gaia-val
+uv run main.py get gaia-val
 ## run the code
 cd miroflow/apps/run-agent
 uv run main.py common-benchmark benchmark=gaia-validation

apps/run-agent/.env.template

@@ -23,4 +23,5 @@ NEWAPI_API_KEY=""
 NEWAPI_BASE_URL=""
 
 # use HTTPS proxy
-HTTPS_PROXY=""
+HTTPS_PROXY=""
+DATA_DIR="../../data"

apps/run-agent/scripts/claude-sonnet-3.7/run_evaluate_multiple_runs_gaia-validation.sh

@@ -25,7 +25,7 @@ for i in $(seq 1 $NUM_RUNS); do
     RUN_ID="run_$i"
 
     (
-        uv run python benchmarks/common_benchmark.py \
+        uv run main.py common-benchmark \
             benchmark=$BENCHMARK_NAME \
             llm=claude_openrouter \
             llm.provider=$LLM_PROVIDER \
@@ -35,7 +35,7 @@ for i in $(seq 1 $NUM_RUNS); do
             benchmark.execution.max_concurrent=5 \
             benchmark.execution.pass_at_k=1 \
             agent=$AGENT_SET \
-            hydra.run.dir=${RESULTS_DIR}/$RUN_ID \
+            output_dir="$RESULTS_DIR/$RUN_ID" \
             > "$RESULTS_DIR/${RUN_ID}_output.log" 2>&1
 
         if [ $? -eq 0 ]; then
@@ -64,7 +64,7 @@ echo "All $NUM_RUNS runs completed!"
 echo "=========================================="
 
 echo "Calculating average scores..."
-uv run python benchmarks/evaluators/calculate_average_score.py "$RESULTS_DIR"
+uv run main.py avg-score "$RESULTS_DIR"
 
 echo "=========================================="
 echo "Multiple runs evaluation completed!"

docs/e2b.Dockerfile

@@ -1,3 +1,14 @@
+# You can use most Debian-based base images
+FROM e2bdev/code-interpreter
+
+# Update package list and install Python 3.10 and pip
+RUN apt-get update && apt-get install -y \
+    portaudio19-dev \
+    && rm -rf /var/lib/apt/lists/*
+
+RUN python3 -m pip install --no-cache-dir --upgrade pip setuptools wheel
+
+# Install dependencies and customize sandbox
 RUN python3 -m pip install --no-cache-dir \
     Flask \
     IPython \

libs/miroflow-tool/src/miroflow/tool/mcp_servers/python_server.py

@@ -1,8 +1,8 @@
+import asyncio
 import os
 
 from e2b_code_interpreter import Sandbox
 from fastmcp import FastMCP
-import asyncio
 
 # Initialize FastMCP server
 mcp = FastMCP("e2b-python-interpreter")
@@ -11,6 +11,10 @@
 E2B_API_KEY = os.environ.get("E2B_API_KEY")
 LOGS_DIR = os.environ.get("LOGS_DIR")  # Directory where benchmark logs are stored
 
+# DEFAULT TEMPLATE ID
+# see README.md on how to build this
+DEFAULT_TEMPLATE_ID = "1av7fdjfvcparqo8efq6"
+
 # DEFAULT CONFS
 DEFAULT_TIMEOUT = 1200  # seconds
 
@@ -138,7 +142,9 @@ async def create_sandbox() -> str:
         sandbox = None
         try:
             sandbox = Sandbox(
-                template="all_pip_apt_pkg", timeout=DEFAULT_TIMEOUT, api_key=E2B_API_KEY
+                template=DEFAULT_TEMPLATE_ID,
+                timeout=DEFAULT_TIMEOUT,
+                api_key=E2B_API_KEY,
             )
             info = sandbox.get_info()