可信 — 风险评分 0/100
上次扫描:1 天前 重新扫描
0 /100
agent-memory-architecture
Implement the 5-file agent memory architecture for durable continuity across sessions.
Pure documentation skill describing agent memory architecture patterns — no executable code, no scripts, no network access, and no sensitive resource usage.
技能名称agent-memory-architecture
分析耗时19.6s
引擎pi
可以安装
No action needed. This skill is safe to use as it contains only markdown documentation.
资源类型声明权限推断权限状态证据
文件系统 NONE NONE SKILL.md only describes file-naming conventions; no file writes are performed by…
网络访问 NONE NONE No network requests anywhere in the skill
命令执行 NONE NONE No shell commands or subprocess calls; mkdir/touch in Section 7 are example outp…
环境变量 NONE NONE No environment variable access in any file
技能调用 NONE NONE No cross-skill invocation observed
剪贴板 NONE NONE No clipboard operations
浏览器 NONE NONE No browser usage
数据库 NONE NONE No database access

目录结构

3 文件 · 41.1 KB · 1163 行
Markdown 3f · 1163L
├─ 📁 references
│ ├─ 📝 memory-templates.md Markdown 183L · 3.2 KB
│ └─ 📝 typing-guide.md Markdown 136L · 4.4 KB
└─ 📝 SKILL.md Markdown 844L · 33.5 KB

安全亮点

✓ Only contains markdown documentation — no executable code whatsoever
✓ No scripts, binaries, or dependency files present
✓ Strong security documentation: explicit MEMORY.md loading boundary (main session only), sacred file definitions, and prompt injection defense rules
✓ No network, shell, filesystem, or credential access declared or implied
✓ References within the skill are local and static (references/memory-templates.md, references/typing-guide.md)
✓ Security guidance is well-designed: read-only default for integrations, WAL protocol for persistence, Level 3 script gates for critical rules