Oynix is an engineering memory layer for teams using AI coding agents. It connects code with tickets, docs, and conversations in a live graph, so agents can retrieve ownership, decisions, and rationale.

What's included in this release:

• Oynix parses repositories into a live code graph, exposing symbols, calls, imports, and relationships.
• Source connectors link code with tickets, documents, conversations, errors, and files so context stays close to the work.
• MCP retrieval gives Claude, Cursor, Codex, Gemini CLI, and other agents graph facts for existing workflows.
• Local-first operation keeps the engine, graph, database, and provider keys on infrastructure the team controls.
• The generated wiki turns connected evidence into an engineering handbook that stays current as code changes.

Together, these capabilities help an engineer move from a code symbol to the decision, owner, or discussion behind it, giving each agent a more useful starting point.