MemSkill Model Card
这是可索引项目报告证据页：它保留 MemSkill Model Card 的源材料入口、机制线索和限制提醒；正文仍需 reader/editor 与 academic public-copy review 后才能当作最终结论引用。
MemSkill Model Card
| Field | Value |
|---|---|
| Repository | ViktorAxelsen/MemSkill |
| Category | Evolving Memory Skills |
| Stars / forks snapshot | 484 / 31 |
| Language | Python |
| License | Apache-2.0 |
| Raw capture | raw-github/viktoraxelsen_memskill.md |
| Updated by | hourly public metadata update, 2026-05-25 |
1. Role in Self Evolve
MemSkill 是学习并进化 long-horizon agent 记忆技能的框架，把记忆操作从静态手写规则变成由任务反馈驱动的 meta-memory skill bank。
2. Working Principle
long-horizon interaction data -> skill-conditioned memory construction -> hard-case mining -> memory skill refinement/new skill proposals -> reusable skill bank
3. Evidence Path
web GitHub page observed 33 commits, Apache-2.0 license, 484 stars and 31 forks; README says MemSkill learns and evolves memory skills for long-horizon agents, replaces static memory operations with a data-driven loop, mines hard cases to refine or propose skills, and evaluates ALFWorld/LoCoMo/LongMemEval style workloads. Shell GitHub API access remained blocked by DNS and local gh auth was invalid in this run, so this card treats the current snapshot as web-observed rather than API-verified.
4. Teaching Use
Use this card to explain Evolving Memory Skills in the raw -> classification -> project card -> site/report pipeline. The reading path is: raw capture -> classification row -> public site card -> project report -> aggregate GitHub analysis.
5. Limits
当前未克隆源码，未运行 benchmark、skill install flows、memory experiments、MCP servers、agent evolution loops、security scanners 或 production deployments；star/fork/commit/release 快照来自公开 GitHub 页面文本或可见页面片段。