https://github.com/zozo123/langchain-islo
Islo sandbox backend for Deep Agents (LangChain) - follows the Daytona partner pattern
https://github.com/zozo123/langchain-islo
Last synced: 11 days ago
JSON representation
Islo sandbox backend for Deep Agents (LangChain) - follows the Daytona partner pattern
- Host: GitHub
- URL: https://github.com/zozo123/langchain-islo
- Owner: zozo123
- License: mit
- Created: 2026-06-05T10:16:53.000Z (29 days ago)
- Default Branch: main
- Last Pushed: 2026-06-05T10:51:29.000Z (29 days ago)
- Last Synced: 2026-06-05T12:26:01.355Z (29 days ago)
- Language: Python
- Size: 18.6 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# langchain-islo
[](https://pypi.org/project/langchain-islo/#history) [](https://opensource.org/licenses/MIT) [](https://pypistats.org/packages/langchain-islo)
Islo sandbox integration for Deep Agents.
Looking for the JS/TS version? Check out the Islo SDK or LangChain.js partners when available.
## Quick Install
```bash
pip install langchain-islo
```
```python
from islo import Islo
from langchain_islo import IsloSandbox
client = Islo()
sandbox = client.sandboxes.create_sandbox(
name="langchain-demo",
image="docker.io/library/ubuntu:24.04",
vcpus=2,
memory_mb=4096,
)
backend = IsloSandbox(client=client, sandbox=sandbox)
result = backend.execute("python --version")
print(result.output)
# Clean up when done
client.sandboxes.delete_sandbox(sandbox_name=sandbox.name)
```
## 🤔 What is this?
Islo sandbox integration for Deep Agents. Islo provides fast, persistent, agent-optimized cloud sandboxes purpose-built for AI coding agents.
See https://islo.dev and the [Islo docs](https://docs.islo.dev) for signup, authentication (ISLO_API_KEY), and platform details.
## 📕 Releases & Versioning
See our [Releases](https://docs.langchain.com/oss/python/release-policy) and [Versioning](https://docs.langchain.com/oss/python/versioning) policies.
## 💁 Contributing
As an open-source project in a rapidly developing field, we are extremely open to contributions, whether it be in the form of a new feature, improved infrastructure, or better documentation.
For detailed information on how to contribute, see the [Contributing Guide](https://docs.langchain.com/oss/python/contributing/overview).
## Usage with Deep Agents
```python
from islo import Islo
from deepagents import create_deep_agent
from langchain_anthropic import ChatAnthropic
from langchain_islo import IsloSandbox
client = Islo()
sandbox = client.sandboxes.create_sandbox(
name="agent-workspace",
vcpus=4,
memory_mb=8192,
)
backend = IsloSandbox(client=client, sandbox=sandbox)
agent = create_deep_agent(
model=ChatAnthropic(model="claude-sonnet-4-6"),
system_prompt="You are a Python coding assistant with sandbox access.",
backend=backend,
)
try:
result = agent.invoke(
{
"messages": [
{
"role": "user",
"content": "Create a small Python package and run pytest",
}
]
}
)
finally:
client.sandboxes.delete_sandbox(sandbox_name=sandbox.name)
```
## File operations (upload / download)
Use the native fast paths:
```python
backend.upload_files(
[
("/src/main.py", b"print('hi from islo')\n"),
("/requirements.txt", b"requests\n"),
]
)
files = backend.download_files(["/src/main.py"])
print(files[0].content)
```
## CLI support (deepagents-code / dcode)
Once published and the provider is wired in deepagents-code (similar to the E2B / langchain-e2b flow), you will be able to do:
```bash
uvx deepagents-cli --sandbox islo ...
# or inside the TUI: /install islo
```
See the E2B sandbox backend issue in langchain-ai/deepagents for the pattern.
## License
MIT
## Integration Status
- Package: https://github.com/zozo123/langchain-islo
- Support PR / Issue in deepagents (LangChain ecosystem): https://github.com/langchain-ai/deepagents/issues/3776 and https://github.com/langchain-ai/deepagents/pull/3775
## Submitting the supporting issue (required process)
To get the PR reviewed, an issue must first be opened **manually via the GitHub web UI** using a template (CLI-created ones get auto-closed).
See `HOW_TO_SUBMIT_ISSUE.md` for exact steps and copy-paste text.
Files prepared:
- `HOW_TO_SUBMIT_ISSUE.md` — full instructions
- `READY_ISSUE_BODY.md` — the body text to paste into the form