RLM-style long-context controller that treats inputs as external context, slices/peeks/searches, and spawns recursive subcalls with strict safety limits. Use for huge docs, dense logs, or repository-scale analysis.

采用RLM风格的长上下文控制器，将输入视为外部上下文，支持切片、窥视和搜索，并在严格的安全限制下生成递归子调用，适用于超大文档、密集日志或仓库规模的分析