langchain-agentcore-codeinterpreter 0.0.5__tar.gz → 0.0.6__tar.gz

{langchain_agentcore_codeinterpreter-0.0.5 → langchain_agentcore_codeinterpreter-0.0.6}/.gitignore

@@ -117,6 +117,11 @@ celerybeat.pid
 # Environments
 .env
 .envrc
+.env.*
+*.pem
+*.key
+*.crt
+credentials.json
 .venv*
 venv*
 env/

{langchain_agentcore_codeinterpreter-0.0.5 → langchain_agentcore_codeinterpreter-0.0.6}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.4
+Metadata-Version: 2.5
 Name: langchain-agentcore-codeinterpreter
-Version: 0.0.5
+Version: 0.0.6
 Summary: Amazon Bedrock AgentCore Code Interpreter sandbox integration for Deep Agents
 Project-URL: Source Code, https://github.com/langchain-ai/langchain-aws/tree/main/libs/agentcore-codeinterpreter
 Project-URL: Repository, https://github.com/langchain-ai/langchain-aws
@@ -8,9 +8,9 @@ License-Expression: MIT
 License-File: LICENSE
 Keywords: agentcore,aws,bedrock,code-interpreter,deepagents,langchain,sandbox
 Requires-Python: <4.0,>=3.11
-Requires-Dist: bedrock-agentcore>=1.1.4
+Requires-Dist: bedrock-agentcore>=1.1.5
 Requires-Dist: boto3>=1.42.42
-Requires-Dist: deepagents>=0.1.0
+Requires-Dist: deepagents>=0.5.6
 Description-Content-Type: text/markdown
 
 # langchain-agentcore-codeinterpreter
@@ -21,7 +21,7 @@ Description-Content-Type: text/markdown
 
 Amazon Bedrock AgentCore Code Interpreter sandbox integration for [Deep Agents](https://github.com/langchain-ai/deepagents).
 
-This package provides `AgentCoreSandbox` — a [`SandboxBackendProtocol`](https://docs.langchain.com/oss/deepagents/sandboxes) implementation that wraps AgentCore's [Code Interpreter](https://docs.aws.amazon.com/bedrock/latest/userguide/agentcore-code-interpreter.html), a secure, isolated MicroVM environment for executing code. The caller manages the interpreter lifecycle (`start()` / `stop()`); the sandbox backend handles command execution and file operations.
+This package provides `AgentCoreSandbox` — a [`SandboxBackendProtocol`](https://docs.langchain.com/oss/deepagents/sandboxes) implementation that wraps AgentCore's [Code Interpreter](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/code-interpreter-tool.html), a secure, isolated MicroVM environment for executing code. `AgentCoreSandbox.create()` starts a session and stops it when you close the sandbox or leave a `with` block.
 
 > **Note:** For the LangChain `BaseTool` integration (used with `create_react_agent` and LangGraph agents), see [`langchain-aws[tools]`](https://github.com/langchain-ai/langchain-aws). This package is specifically for the Deep Agents `BaseSandbox` protocol.
 
@@ -50,9 +50,9 @@ aws sso login --profile your-profile
 
 Any method supported by the [boto3 credential chain](https://boto3.amazonaws.com/v1/documentation/api/latest/guide/credentials.html) works, including EC2 instance profiles, ECS task roles, and environment variables.
 
-**2. IAM permissions** — your credentials must allow `bedrock-agentcore:InvokeCodeInterpreter` (or the equivalent action for your region). See the [AgentCore Code Interpreter docs](https://docs.aws.amazon.com/bedrock/latest/userguide/agentcore-code-interpreter.html) for the required IAM policy.
+**2. IAM permissions** — your credentials must allow `bedrock-agentcore:InvokeCodeInterpreter` (or the equivalent action for your region). See the [AgentCore Code Interpreter docs](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/code-interpreter-tool.html) for the required IAM policy.
 
-**3. Region availability** — Code Interpreter is available in select AWS regions. `us-west-2` is a safe default. Pass the region to `CodeInterpreter(region=...)`.
+**3. Region availability** — Code Interpreter is available in select AWS regions. `us-west-2` is a safe default. Pass the region to `AgentCoreSandbox.create(region=...)`.
 
 ## Quick Install
 
@@ -65,45 +65,52 @@ pip install langchain-agentcore-codeinterpreter
 ### Standalone
 
 ```python
-from bedrock_agentcore.tools.code_interpreter_client import CodeInterpreter
-
 from langchain_agentcore_codeinterpreter import AgentCoreSandbox
 
-interpreter = CodeInterpreter(region="us-west-2")
-interpreter.start()
+with AgentCoreSandbox.create(region="us-west-2") as backend:
+    result = backend.execute("echo hello")
+    print(result.output, result.exit_code)  # "hello" 0
+```
+
+Sessions last 15 minutes by default. Pass `session_timeout_seconds` (up to 28,800, or 8 hours) for longer agent runs, and `identifier` to use a custom code interpreter with VPC networking or an execution role.
 
-backend = AgentCoreSandbox(interpreter=interpreter)
+To configure the `CodeInterpreter` client yourself, pass a started one to the constructor instead. You then own its lifecycle:
 
-result = backend.execute("echo hello")
-print(result.output)  # "hello"
+```python
+from bedrock_agentcore.tools.code_interpreter_client import CodeInterpreter
 
-interpreter.stop()
+interpreter = CodeInterpreter(
+    region="us-west-2", integration_source="langchain-deepagents"
+)
+interpreter.start()
+try:
+    backend = AgentCoreSandbox(interpreter=interpreter)
+    ...
+finally:
+    interpreter.stop()
 ```
 
 ### With Deep Agents
 
 ```python
-from bedrock_agentcore.tools.code_interpreter_client import CodeInterpreter
 from deepagents import create_deep_agent
 
 from langchain_agentcore_codeinterpreter import AgentCoreSandbox
 from langchain_aws import ChatBedrockConverse
 
-interpreter = CodeInterpreter(region="us-west-2")
-interpreter.start()
-
 model = ChatBedrockConverse(
-  model="us.anthropic.claude-sonnet-4-6",
-  region_name="us-west-2",
-)
-backend = AgentCoreSandbox(interpreter=interpreter)
-agent = create_deep_agent(
-    model=model,
-    backend=backend,
-    system_prompt="You are a coding assistant with sandbox access.",
+    model="us.anthropic.claude-sonnet-4-6",
+    region_name="us-west-2",
 )
 
-try:
+with AgentCoreSandbox.create(
+    region="us-west-2", session_timeout_seconds=3600
+) as backend:
+    agent = create_deep_agent(
+        model=model,
+        backend=backend,
+        system_prompt="You are a coding assistant with sandbox access.",
+    )
     result = agent.invoke(
         {
             "messages": [
@@ -112,18 +119,18 @@ try:
         }
     )
     print(result["messages"][-1].content)
-finally:
-    interpreter.stop()
 ```
 
 ### File operations
 
 ```python
 # Upload files
-backend.upload_files([
-    ("data.csv", b"name,value\nalice,42\nbob,17"),
-    ("analyze.py", b"import csv\nprint('ready')"),
-])
+backend.upload_files(
+    [
+        ("data.csv", b"name,value\nalice,42\nbob,17"),
+        ("analyze.py", b"import csv\nprint('ready')"),
+    ]
+)
 
 # Download files
 results = backend.download_files(["data.csv"])
@@ -136,7 +143,11 @@ for r in results:
 
 ## Session behavior
 
-AgentCore sessions cannot be reconnected after `interpreter.stop()` is called. Each `start()` creates a fresh, isolated MicroVM. Sessions auto-expire after a configurable timeout (default 15 minutes, maximum 8 hours).
+AgentCore sessions cannot be reconnected once stopped. Each session is a fresh, isolated MicroVM, and its files are gone when it ends. Sessions auto-expire after `session_timeout_seconds` (default 15 minutes, maximum 8 hours); calls after that return a session-expired error.
+
+`execute()` honors `timeout` by running the command under GNU `timeout` inside the sandbox. A command stopped this way returns exit code 124.
+
+AgentCore caps each request and response at 100 MB and gives each session 10 GB of disk. See [AgentCore quotas](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/bedrock-agentcore-limits.html).
 
 ## Contributing
 

{langchain_agentcore_codeinterpreter-0.0.5 → langchain_agentcore_codeinterpreter-0.0.6}/README.md

@@ -6,7 +6,7 @@
 
 Amazon Bedrock AgentCore Code Interpreter sandbox integration for [Deep Agents](https://github.com/langchain-ai/deepagents).
 
-This package provides `AgentCoreSandbox` — a [`SandboxBackendProtocol`](https://docs.langchain.com/oss/deepagents/sandboxes) implementation that wraps AgentCore's [Code Interpreter](https://docs.aws.amazon.com/bedrock/latest/userguide/agentcore-code-interpreter.html), a secure, isolated MicroVM environment for executing code. The caller manages the interpreter lifecycle (`start()` / `stop()`); the sandbox backend handles command execution and file operations.
+This package provides `AgentCoreSandbox` — a [`SandboxBackendProtocol`](https://docs.langchain.com/oss/deepagents/sandboxes) implementation that wraps AgentCore's [Code Interpreter](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/code-interpreter-tool.html), a secure, isolated MicroVM environment for executing code. `AgentCoreSandbox.create()` starts a session and stops it when you close the sandbox or leave a `with` block.
 
 > **Note:** For the LangChain `BaseTool` integration (used with `create_react_agent` and LangGraph agents), see [`langchain-aws[tools]`](https://github.com/langchain-ai/langchain-aws). This package is specifically for the Deep Agents `BaseSandbox` protocol.
 
@@ -35,9 +35,9 @@ aws sso login --profile your-profile
 
 Any method supported by the [boto3 credential chain](https://boto3.amazonaws.com/v1/documentation/api/latest/guide/credentials.html) works, including EC2 instance profiles, ECS task roles, and environment variables.
 
-**2. IAM permissions** — your credentials must allow `bedrock-agentcore:InvokeCodeInterpreter` (or the equivalent action for your region). See the [AgentCore Code Interpreter docs](https://docs.aws.amazon.com/bedrock/latest/userguide/agentcore-code-interpreter.html) for the required IAM policy.
+**2. IAM permissions** — your credentials must allow `bedrock-agentcore:InvokeCodeInterpreter` (or the equivalent action for your region). See the [AgentCore Code Interpreter docs](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/code-interpreter-tool.html) for the required IAM policy.
 
-**3. Region availability** — Code Interpreter is available in select AWS regions. `us-west-2` is a safe default. Pass the region to `CodeInterpreter(region=...)`.
+**3. Region availability** — Code Interpreter is available in select AWS regions. `us-west-2` is a safe default. Pass the region to `AgentCoreSandbox.create(region=...)`.
 
 ## Quick Install
 
@@ -50,45 +50,52 @@ pip install langchain-agentcore-codeinterpreter
 ### Standalone
 
 ```python
-from bedrock_agentcore.tools.code_interpreter_client import CodeInterpreter
-
 from langchain_agentcore_codeinterpreter import AgentCoreSandbox
 
-interpreter = CodeInterpreter(region="us-west-2")
-interpreter.start()
+with AgentCoreSandbox.create(region="us-west-2") as backend:
+    result = backend.execute("echo hello")
+    print(result.output, result.exit_code)  # "hello" 0
+```
+
+Sessions last 15 minutes by default. Pass `session_timeout_seconds` (up to 28,800, or 8 hours) for longer agent runs, and `identifier` to use a custom code interpreter with VPC networking or an execution role.
 
-backend = AgentCoreSandbox(interpreter=interpreter)
+To configure the `CodeInterpreter` client yourself, pass a started one to the constructor instead. You then own its lifecycle:
 
-result = backend.execute("echo hello")
-print(result.output)  # "hello"
+```python
+from bedrock_agentcore.tools.code_interpreter_client import CodeInterpreter
 
-interpreter.stop()
+interpreter = CodeInterpreter(
+    region="us-west-2", integration_source="langchain-deepagents"
+)
+interpreter.start()
+try:
+    backend = AgentCoreSandbox(interpreter=interpreter)
+    ...
+finally:
+    interpreter.stop()
 ```
 
 ### With Deep Agents
 
 ```python
-from bedrock_agentcore.tools.code_interpreter_client import CodeInterpreter
 from deepagents import create_deep_agent
 
 from langchain_agentcore_codeinterpreter import AgentCoreSandbox
 from langchain_aws import ChatBedrockConverse
 
-interpreter = CodeInterpreter(region="us-west-2")
-interpreter.start()
-
 model = ChatBedrockConverse(
-  model="us.anthropic.claude-sonnet-4-6",
-  region_name="us-west-2",
-)
-backend = AgentCoreSandbox(interpreter=interpreter)
-agent = create_deep_agent(
-    model=model,
-    backend=backend,
-    system_prompt="You are a coding assistant with sandbox access.",
+    model="us.anthropic.claude-sonnet-4-6",
+    region_name="us-west-2",
 )
 
-try:
+with AgentCoreSandbox.create(
+    region="us-west-2", session_timeout_seconds=3600
+) as backend:
+    agent = create_deep_agent(
+        model=model,
+        backend=backend,
+        system_prompt="You are a coding assistant with sandbox access.",
+    )
     result = agent.invoke(
         {
             "messages": [
@@ -97,18 +104,18 @@ try:
         }
     )
     print(result["messages"][-1].content)
-finally:
-    interpreter.stop()
 ```
 
 ### File operations
 
 ```python
 # Upload files
-backend.upload_files([
-    ("data.csv", b"name,value\nalice,42\nbob,17"),
-    ("analyze.py", b"import csv\nprint('ready')"),
-])
+backend.upload_files(
+    [
+        ("data.csv", b"name,value\nalice,42\nbob,17"),
+        ("analyze.py", b"import csv\nprint('ready')"),
+    ]
+)
 
 # Download files
 results = backend.download_files(["data.csv"])
@@ -121,7 +128,11 @@ for r in results:
 
 ## Session behavior
 
-AgentCore sessions cannot be reconnected after `interpreter.stop()` is called. Each `start()` creates a fresh, isolated MicroVM. Sessions auto-expire after a configurable timeout (default 15 minutes, maximum 8 hours).
+AgentCore sessions cannot be reconnected once stopped. Each session is a fresh, isolated MicroVM, and its files are gone when it ends. Sessions auto-expire after `session_timeout_seconds` (default 15 minutes, maximum 8 hours); calls after that return a session-expired error.
+
+`execute()` honors `timeout` by running the command under GNU `timeout` inside the sandbox. A command stopped this way returns exit code 124.
+
+AgentCore caps each request and response at 100 MB and gives each session 10 GB of disk. See [AgentCore quotas](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/bedrock-agentcore-limits.html).
 
 ## Contributing