python-1.47.0
Semantic Kernel Python 1.47.0 drops Python 3.10 support (now requiring 3.11-3.14), fixes agent history handling, and introduces several connector and MCP bug fixes.
Why this signal matters
Semantic Kernel release python-1.47.0 drops support for Python 3.10, requiring Python 3.11 or later with tested support extending through Python 3.14. Key Python fixes include integrating ChatHistoryReducer into ChatCompletionAgent invocations, preserving non-adjacent tool pairs during history truncation, handling empty tool and function results, and respecting optional MCP prompt arguments while propagating MCP initialization failures. Additional updates include batching Google AI embedding requests, counting Gemini thinking tokens, handling Redis API changes (redisvl = 0.5), and fixing SQL Server and OracleCollection connector query behavior.
Actionable summary
Semantic Kernel Python 1.47.0 introduces a breaking runtime change by dropping Python 3.10 support in favor of Python 3.11+. It also fixes MCP initialization propagation, agent invocation history truncation, and connector integrations across Redis, SQL Server, and Google AI.
- Agent usefulness
- 89/100
- Confidence
- 96%
- Canonical data
- JSON + Markdown
What builders should check
- Upgrade Python runtime environments to Python 3.11 or later before upgrading to Semantic Kernel Python 1.47.0.
- Verify tool-call and agent invocation logic if relying on ChatCompletionAgent history truncation or empty tool responses.
- Review MCP client initialization handling to ensure propagated initialization failures are caught properly.