Personal agents tailor replies by drawing on past conversations. Their memory systems store and update each user’s utterances, and coding agents now even rewrite the harness that injects this memory into the model’s context. Such patches risk overfitting, fixing only the failures they target. Reglos is a training‑free personal memory agent that tests whether harness changes help beyond those specific failures. Its memory is append‑only: every conversation is added, newer preferences overwrite older ones, and facts can be forgotten. The harness improves via recursive self‑improvement: a coding agent proposes patches from failure traces, and a critic from a second model family screens out potential benchmark gaming. After the harness is frozen, a fixed evaluator decides which further patches to accept. With the same backbone, Reglos reaches 61.3%/84.0% on PersonaMem‑v2 and LongMemEval‑S with Qwen3.8‑27B, and 57.5%/87.2% with GPT‑5.6 Luna, surpassing state‑of‑the‑art systems such as Mem0, MemOS, and EverOS.\n\nReview