菜单
Initialize an AgentDB Cognitive Container (.rvf file) in the current project. Sets up storage, embedder config, and the agentdb MCP server. Use when the user is starting a new project that needs vector memory, or asks to "set up agentdb" / "init agentdb".
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/ruvnet/agentdb --skill agentdb-init命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Walk the causal graph in AgentDB to explain why two memories are connected, or trace a root cause. Use when the user asks "why did X happen", "what led to Y", or after an incident.
Record a causal relationship between two memories in AgentDB — "X caused Y", "A supersedes B", "patch-foo depends-on patch-bar". Use when the user is documenting cause/effect, dependencies, supersessions, or after-action analysis.
Store memories in tier-aware hierarchical memory — working / short-term / long-term — and recall with tier filters. Use for working-set context that should fade, vs facts that should persist, vs patterns that should be searchable forever.
| name | agentdb-init |
| description | Initialize an AgentDB Cognitive Container (.rvf file) in the current project. Sets up storage, embedder config, and the agentdb MCP server. Use when the user is starting a new project that needs vector memory, or asks to "set up agentdb" / "init agentdb". |
Sets up a fresh AgentDB instance for the current project. Creates a single-file .rvf Cognitive Container that holds vectors, indexes, learning state, and the audit log.
./memory.rvf at the project root, or ~/.agentdb/<project-name>.rvf for global memory).Xenova/all-MiniLM-L6-v2 at 384d — fast, free, runs in-process).claude mcp add agentdb -- npx agentdb@latest mcp start
agentdb_pattern_store (the first store auto-creates the schema), or via CLI:
npx agentdb@latest init ./memory.rvf
*.rvf to .gitignore unless the user explicitly wants memory checked into source control..rvf is a single binary file. Back it up like a SQLite database..rvf per coordinated namespace; use separate files for trust-boundary isolation.agentdb_pattern_*, agentdb_reflexion_*, etc.) or the npm library (import { SelfLearningRvfBackend } from 'agentdb')..rvf file by default — it can hold session-specific data, including content from messages.init on an existing .rvf file without confirming — it will refuse rather than overwrite, but a confused user might delete the existing file thinking it's stale.