Build agents once. Run them anywhere.
KsADK is the Agent Runtime Platform for AI agents. Keep building with Google ADK, LangGraph, LangChain, or DeepAgents, then run, debug, expose, sandbox, deploy, and observe them through one runtime experience.
python -m venv .venv
source .venv/bin/activate
pip install -U "ksadk[all]"
agentengine init demo-agent -f langgraph
cd demo-agent
agentengine config set OPENAI_API_KEY=your-api-key OPENAI_MODEL_NAME=gpt-4o-mini
agentengine run -iStart the local debugging Web UI:
agentengine web . --no-openMost agent frameworks solve how to build agents. KsADK solves how to run, debug, deploy, and observe them.
- Local development:
agentengine init,agentengine run,agentengine web. - Unified debugging: browser Web UI, streaming, attachments, workspace files, tool calls, and sessions.
- Unified protocol: local
/v1/responsesand/v1/chat/completions. - Tool boundaries: Skill Runtime, Workspace, Sandbox, Memory, Knowledge.
- Engineering workflow: packaging, deployment, OpenTelemetry observability.
- Documentation: https://kingsoftcloud.github.io/ksadk-python/
- Quick Start: https://kingsoftcloud.github.io/ksadk-python/en/getting-started/quickstart/
- Why KsADK: https://kingsoftcloud.github.io/ksadk-python/en/getting-started/why-ksadk/
- Architecture: https://kingsoftcloud.github.io/ksadk-python/en/getting-started/architecture/
- Ecosystem Positioning: https://kingsoftcloud.github.io/ksadk-python/en/getting-started/comparison/
- Observability: https://kingsoftcloud.github.io/ksadk-python/en/guides/observability-tracing/
- Samples: https://github.com/kingsoftcloud/ksadk-samples
- KsADK repository: https://github.com/kingsoftcloud/ksadk-python
- Web UI repository: https://github.com/kingsoftcloud/ksadk-web
- Wiki: https://zread.ai/kingsoftcloud/ksadk-python
- PyPI: https://pypi.org/project/ksadk/
Issues, pull requests, samples, and documentation improvements are welcome. Before submitting, run:
make public-preflightLicense: Apache-2.0.



