Existing memory systems for long-horizon LLM agents often retrieve prior traces as passive context rather than converting them into executable capabilities. In this paper, we propose MSCE, a training-free Memory–Skill Co-Evolution framework that organizes agent experience into grounded step traces, reusable procedural policies, and declarative environmental cognition. MSCE crystallizes evidence-backed L2 policies with positive estimated gain into callable skills that retain evidence links, applicability boundaries, decision guidance, verification rules, and reliability estimates. It further introduces reflection-weighted value backfilling, which propagates sparse terminal feedback through dense local self-reflections to produce evidence-calibrated trace values for governing memory and skill evolution. Experiments on EvoAgentBench and LoCoMo show that MSCE achieves the best aggregate performance among the evaluated skill-augmented and memory-driven baselines, with statistically supported aggregate gains, positive cross-domain transfer, and continual improvement as online experience accumulates. Our code is available at https://github.com/MemTensor/MemOS/tree/main/apps/memos-local-plugin.