Fetching from the wire…
Infra2026-09-03 · source-backed
Announced September 2, every LazyFrame query now runs on the streaming engine by default, which the team claims is roughly 5x faster with substantially lower memory on typical workloads. The breaking change with teeth: streaming no longer guarantees row ordering for joins and group_by, so callers depending on it must pass maintain_order=True or fall back to the in-memory engine. is_in now rejects lossy Int64-to-Float64 coercion and horizontal concat requires matching heights. pip install polars==2.0rc1.
Each link below shares sources, entities, or timing with this story.
(intermediate) — Audit for eval/exec/subprocess shell=True, replace with safe alternatives (ast.literal_eval, JSON parsers, execFile + arrays), implement allowlist validation, enable auth. (DEV Community / Kai Security)
Released September 1, content that previously showed completed may now show failed or partial, because those ingests were always incomplete (GitHub). Embedders raise EmbeddingError instead of returning an empty vector, a new ContentStatus.PARTIAL covers files where some chunks...
MCPConfig.tools now accepts Agent, Team and Workflow instances plus Toolkit objects, exposing each as its own named MCP tool. You call chief, not run_agent(agent_id="chief"). Toolkits publish one MCP tool per registered method, narrowed by the toolkit's own enable/include/excl...
Released August 24, an unexpected exception from a tool, resource or prompt handler is now logged once at ERROR locally and the client sees only Error executing tool <name> (GitHub). Anything you intended the model to read must be raised as ToolError or ResourceError. Separate...
Runs now get their own agno_runs table with real columns instead of being packed into the session row, which takes session write amplification from O(N²) to O(N) and removes the DynamoDB and Firestore item-size ceiling (GitHub). offload_tool_results=True writes any tool result...
DeepSeek-V4-Flash-0731 landed July 31 under MIT with a DSpark speculative-decoding module attached. Terminal Bench 2.1: 82.7. Toolathlon-Verified: 70.3. DSBench-FullStack: 68.7. DeepSWE: 54.4. NL2Repo: 54.2. The model card claims it beats DeepSeek-V4-Pro (Preview) "despite its...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.