Getting Started
Requirements
- Python 3.11
- Poetry 1.8.2
- Access to an LLM provider (OpenAI or AWS Bedrock)
- Make sure public SSH key (e.g.
~/.ssh/id_rsa.pub) is added to GitHub (access to repo download) giInstallation
# Clone the repository
git clone git@github.com:stxnext/deep-next.git
cd deep-next
# Install dependencies
make install_venv
# Test the installation
make test_unit
# .env
OPENAI_API_KEY=<your api key>
# Make sure LLM setup is configured properly
# echo "OPENAI_API_KEY=<your api key>" >> .env
make test_core_llm
Example CLI Usage (core pipeline)
git status
poetry run python libs/core/deep_next/core/entrypoint.py \
--problem-statement "Add hello-world feature to core pipeline CLI so that it will print greetings. Make the CLI ok with passing only the hello-world flag. Add feature description to README." \
--hints "..." \
--root-dir "$(pwd)"
git status
App run (GitHub connected)
# This is .env
# OpenAI support (required)
OPENAI_API_KEY=<your api key>
# VCS connector setup (required to connect core pipeline with VCS)
VCS_PROVIDER=github
VCS_ACCESS_TOKEN=<your Github access token>
VCS_REPO_PATH=stxnext/deep-next
Note: Selected scopes for GitHub: project, repo, user, write:discussion
poetry run python apps/app/deep_next/app/entrypoint_scheduled.py
Observability (LangSmith | Langfuse) and Slack notifications
Few additional treats configured via .env
# Langsmith observability (optional)
LANGSMITH_TRACING=true
LANGSMITH_ENDPOINT="https://api.smith.langchain.com"
LANGSMITH_API_KEY="<your api key>"
LANGSMITH_PROJECT="DeepNext"
# Langfuse observability (optional)
LANGFUSE_HOST="https://cloud.langfuse.com"
LANGFUSE_PUBLIC_KEY="<your api key>"
# Slack notifications from app (optional)
SLACK_NOTIFICATIONS=true
SLACK_BOT_TOKEN="<your bot token>"
SLACK_CHANNEL="deepnext-ops"