Memvid Model Card
这是可索引项目报告证据页：它保留 Memvid 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。
Memvid Model Card
| Field | Value |
|---|---|
| Repository | memvid/memvid |
| Category | Single-File Agent Memory Layer |
| Stars / forks snapshot | 12400 / 1000 |
| Language | Python |
| License | Unknown |
| Raw capture | raw-github/memvid_memvid.md |
| Updated by | hourly public metadata update, 2026-05-25 |
1. Role in Self Evolve
Memvid packages AI-agent memory into a portable single-file layer, positioning itself as memory without external vector database/server infrastructure and publishing LoCoMo benchmark claims.
2. Working Principle
documents/conversations -> single portable memory file -> fast retrieval
3. Evidence Path
web search observed memvid/memvid through public GitHub-adjacent and project pages as a single-file portable memory layer for AI agents; indexed snapshot reported 12.4k stars and 1.0k forks, and project benchmark page claims LoCoMo performance gains. Treat benchmark and count signals as web-observed, not API-verified. Shell GitHub API access remained blocked by DNS and local gh auth was invalid, so this card treats the current snapshot as web-observed rather than API-verified.
4. Teaching Use
Use this card to explain Single-File Agent Memory Layer in the raw -> classification -> project card -> site/report pipeline. Does a portable memory artifact improve agent continuity under real coding workloads, not just retrieval benchmarks?
5. Limits
The repository was not cloned in this iteration; no benchmark, SDK example, memory experiment, eval run, skill install flow, agent loop, or production deployment was executed. Counts and claims are visible public-page/search signals unless independently revalidated later.