Releases: derisk-ai/OpenDerisk
Releases · derisk-ai/OpenDerisk
OpenDerisk v0.2.1 Archive
- Archive the latest main branch.
- The latest main archive, version v0.3, is entirely written in Claude code and is incompatible with previous versions.
OpenDerisk v0.2.0
New Features
OpenDerisk V0.2 version builds a future-oriented Multi-Agent development and operation product framework, which mainly includes core capabilities:
- A multi-agent framework with key features including, 1. prompt config, 2. skill configuration (tools/MCP), 3. knowledge, 4. memory, and debugging preview.
- Agentc Workspace for running state, constructing a standard paradigm for multi-agent operation
- New Home page
- Add default scenes
- Troubleshooting expert AI-SRE
- Data analysis expert DataExpert
- Flame graph assistant
For detailed operation manual, please refer to OpenDerisk V0.2 ReleaseNote
What's Changed
- feat: add claude llms and dynamic model choose by @csunny in #14
- fix: remove extra unuseful code by @csunny in #16
- fix: optimize desc and ignore file by @csunny in #17
- style: fix markdown symbols by @anotherso1a in #18
- Reduce the problem of undefined global variables by @dusx1981 in #23
- feat: add tongyi proxy model by @csunny in #25
- feat: add roadmap for openderisk by @csunny in #26
- feat:add storage client cache by @Aries-ckt in #27
- feat: fix mysql table schema error and add aisre eval script by @csunny in #28
- Feature mcp web by @heyzcat in #33
- fix: the parameter passing bug when connecting to mcp by @gallopxiong in #43
- Dev/open derisk v0.2.2 by @yhjun1026 in #47
- feat: open derisk clean temp file by @yhjun1026 in #49
- Dev/fix default model by @yhjun1026 in #50
- docs: add docs for how to use by @csunny in #51
- feat: optimize the dialogue entry code logic by @yhjun1026 in #52
- feat: chat bug fixed by @XinyueDu in #53
- feat: workflow node-version change by @XinyueDu in #54
- feat: fixed stop chat by @XinyueDu in #55
- feat: web theme & prompt edit fixed by @XinyueDu in #56
- feat: mcp by @XinyueDu in #61
- feat: add new vis components config file by @XinyueDu in #63
- docs: add Japanese README by @eltociear in #64
- docs: add deepwiki for OpenDerisk v0.2 by @csunny in #65
- feat: fixed Agent update loading by @XinyueDu in #66
- fix:batch process embeddings while call OpenAI embedding API by @lpq131004 in #68
- feat: Website i18n Translation by @XinyueDu in #69
- feat: add-knowledge-chunk by @XinyueDu in #70
- feat: add mcp server registry and call features by @csunny in #71
- feat: add mermaid by @XinyueDu in #72
- feat: build mcp frontend & fix agents translate by @csunny in #73
- Feat/fix kn by @yhjun1026 in #74
- Feat/fix kn by @yhjun1026 in #75
- fix:reasoning agent find knowledge empty. by @Aries-ckt in #76
- feat: add mcp by @XinyueDu in #78
- docs: add docs for OpenDerisk v0.2 version by @csunny in #80
New Contributors
- @anotherso1a made their first contribution in #18
- @dusx1981 made their first contribution in #23
- @Aries-ckt made their first contribution in #27
- @heyzcat made their first contribution in #33
- @gallopxiong made their first contribution in #43
- @yhjun1026 made their first contribution in #47
- @eltociear made their first contribution in #64
- @lpq131004 made their first contribution in #68
Full Changelog: v0.1.0...v0.2.0
v0.1.0
OpenDerisk v0.1.0
OpenDerisk — your intelligent application risk manager — delivers 7*24H comprehensive system protection. As the first fully open-source AI-driven SRE multi-agent project, it pioneers intelligent risk governance.
Core features
- DeepResearch RCA: Quickly locate the root cause of issues through in-depth analysis of logs, traces, and code.
- Multi-Agent Collaboration: Collaboration among SRE-Agent, Code-Agent, ReportAgent, Vis-Agent, and Data-Agent.
- Open and Open-Source Architecture: OpenDerisk is built with a completely open and open-source architecture, allowing related frameworks and code to be used out of the box in open-source projects.
- Visualized Evidence Chain: Fully visualize the diagnostic process and evidence chain, making the diagnosis clear and enabling quick judgment of accuracy.
This is the first version we've released, and we're building something amazing - stay tuned for exciting updates! 👏👏👏🎉🎉🎉