OpenDerisk
AI-Native Risk Intelligence Systems, OpenDeRisk——Your application system risk intelligent manager provides 7* 24-hour comprehensive and in-depth protection.
claude mcp add --transport stdio derisk-ai-openderisk uvx run python packages/derisk-app/src/derisk_server.py --config configs/derisk-proxy-aliyun.toml \ --env DERISK_API_KEY="Your OpenDeRisk API key" \ --env DERISK_CONFIG_PATH="Path to derisk-proxy-aliyun.toml if applicable" \ --env DERISK_DATASETS_DIR="pilot/datasets (default location for OpenRCA dataset)"
How to use
OpenDerisk is an AI-native risk intelligence system that coordinates multiple agents and tools to perform root cause analysis, evidence visualization, and data-driven risk assessment. The server is intended to run under uv and serves a web UI for interacting with the OpenRCA-powered diagnostics, Flame Graph Assistant, and DataExpert-style data ingestion capabilities. Once running, you can access the web UI at http://localhost:7777 to view the diagnostics pipeline, manage datasets, and review evidence chains generated by the multi-agent system. Use the provided configuration to point OpenDerisk at your API keys and dataset locations, enabling automated observation and reporting.
How to install
Prerequisites
- Python 3.8+ and uv (the Python-based server runner) installed on your system
- Git installed
- Sufficient disk space for the OpenRCA dataset (~26GB decompressed, plus processing space)
Installation steps
- Clone the repository
git clone https://github.com/derisk-ai/OpenDerisk.git
cd OpenDerisk
- Prepare the release environment (needed for development setup)
sh scripts/prepare_release.sh
- Install dependencies for development (via the recommended uv workflow)
# Ensure you are in the repository root
uv install
- Download the default dataset (Telecom/OpenRCA dataset) if you plan to reproduce the default setup
gdown https://drive.google.com/uc?id=1cyOKpqyAP4fy-QiJ6a_cKuwR7D46zyVe
- Place datasets as indicated by the guide
mv downloaded_dataset_path pilot/datasets/
- Run the server using the provided configuration
uv run python packages/derisk-app/src/derisk_server.py --config configs/derisk-proxy-aliyun.toml
- Open the web UI at http://localhost:7777
Additional notes
Notes and tips
- The system relies on the OpenRCA dataset for root cause analysis. Ensure you have sufficient disk space (the decompressed dataset is large, ~26GB).
- The configuration file derisk-proxy-aliyun.toml must be created and filled with your API keys and any required dataset paths. The API key usually comes from your OpenDeRisk deployment credentials.
- If you encounter port conflicts, you can modify the configuration or run the server on a different port if supported by your setup.
- For AI-SRE workflows, you can use the AI-SRE OpenRCA mode to perform root cause analysis on logs, traces, and metrics; the Flame Graph Assistant can ingest flame graphs for performance diagnostics; and the DataExpert tool can ingest metrics/logs/traces for conversational analysis.
- Ensure network access to required data sources and OpenRCA endpoints, and keep the dataset paths consistent with the configuration file.
Related MCP Servers
FastGPT
FastGPT is a knowledge-based platform built on the LLMs, offers a comprehensive suite of out-of-the-box capabilities such as data processing, RAG retrieval, and visual AI workflow orchestration, letting you easily develop and deploy complex question-answering systems without the need for extensive setup or configuration.
note-gen
A cross-platform Markdown AI note-taking software.
Upsonic
Agent Framework For Fintech and Banks
shellfirm
Terminal guardrails for humans and AI agents — intercepts dangerous commands with context-aware challenges before the damage is done.
AEnvironment
Standardized environment infrastructure for Agentic AI development.
omnicoreagent
OmniCoreAgent is a powerful Python framework for building autonomous AI agents that think, reason, and execute complex tasks. Production-ready agents that use tools, manage memory, coordinate workflows, and handle real-world business logic.
