agentmods.dev

RAG快照机制完善需求文档

A Chinese-language requirements document for completing an RAG snapshot system. RAG, or retrieval-augmented generation, lets an AI answer using a connected knowledge base; a snapshot is a fixed copy of that data.

In plain words
What is it for?
Use it as a development reference for RAG versioning, file and document snapshots, vector-data isolation, publishing and installation APIs, and related Java modules.
Why use it?
It describes how editable source data, published versions, and users' installed copies should be separated. It also identifies missing snapshot data, confusing access rules, and unfinished service methods.

Command for Claude Code

830 +1 2mo ago A scan 0 tokens original Apache-2.0
View the source file lucky-aeon/AgentX
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/lucky-aeon/agentx/rag
Clone the repo
git clone --depth 1 https://github.com/lucky-aeon/AgentX

Made for: Claude Code.

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,099 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.02099
Opus 5 $0.00000 $0.01050
Sonnet 5 $0.00000 $0.00420
Haiku 4.5 $0.00000 $0.00210

Measured 2d ago against content hash 017bb1e29057, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

RAG快照机制完善需求文档 scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/commands/RAG快照机制完善需求文档.md · 204 lines

How it starts

The opening of the file, as written. The whole thing — 204 lines — stays where its author put it; the contents beside it link to each section on GitHub.

RAG快照机制完善需求文档

业务背景

RAG系统架构概述

AgentX的RAG（知识库）系统采用分层发布机制：

  1. 原始RAG数据集：用户创建和维护的工作数据，可随时修改
  2. 版本快照：发布到市场的固化版本，内容不可变
  3. 用户安装快照：用户从市场安装的个人副本，完全隔离

版本管理机制

  • 用户创建RAG后自动生成0.0.1版本（REFERENCE类型），仅创建者可见
  • 用户可发布RAG版本到市场（1.0.0、1.1.0等），生成SNAPSHOT类型的快照
  • 其他用户只能看到和安装已发布的SNAPSHOT版本
  • 0.0.1版本永远不会被其他用户安装

安装类型说明

REFERENCE类型（引用类型）：

  • 仅用于用户自己创建的RAG的0.0.1版本
  • 动态引用原始RAG数据，随原始数据变化而更新
  • 用户修改原始RAG时，对应的0.0.1版本安装记录也会同步更新

SNAPSHOT类型（快照类型）：

  • 用于所有发布版本和他人安装的版本
  • 完全独立的数据副本，与原始RAG隔离
  • 安装后内容固定，不会因原始RAG变化而改变

当前问题分析

1. 快照数据不完整

目前的快照机制只复制了基本信息（name、description、icon），但缺少：

  • 文件数据的完整快照
  • 文档单元的完整快照
  • 相关配置和元数据的快照

2. 数据访问逻辑混乱

在查看已安装RAG信息时：

  • REFERENCE类型应该显示原始RAG的实时信息
  • SNAPSHOT类型应该显示快照数据
  • 但当前实现中都从rag_versions表获取信息，违背了快照原则

3. 快照服务未完善

RagDataAccessService中的快照方法大多返回空列表，未实现真正的快照数据获取。

需要修改的代码模块

1. 数据模型扩展

UserRagDTO (/AgentX/src/main/java/org/xhy/application/rag/dto/UserRagDTO.java)

修改内容：

  • 添加installType字段
  • 添加isReferenceType()和isSnapshotType()判断方法
UserRagEntity 相关表结构

确认字段：

  • install_type字段已存在
  • 快照数据字段（name、description、icon等）已存在

2. 快照创建机制

RagPublishAppService (/AgentX/src/main/java/org/xhy/application/rag/RagPublishAppService.java)

需要完善的功能：

  • 发布版本时创建完整快照
  • 复制所有文件到rag_version_files表
  • 复制所有文档单元到rag_version_documents表
  • 复制相关配置和元数据
RagVersionDomainService (/AgentX/src/main/java/org/xhy/domain/rag/service/RagVersionDomainService.java)

需要添加的方法：

  • createCompleteSnapshot() - 创建完整版本快照
  • copyFilesToVersion() - 复制文件快照
  • copyDocumentsToVersion() - 复制文档快照

3. 快照安装机制

UserRagDomainService (/AgentX/src/main/java/org/xhy/domain/rag/service/UserRagDomainService.java)

需要完善的方法：

  • installRag() - 安装SNAPSHOT类型时创建用户专属快照
  • 添加快照数据复制逻辑
新增服务类建议

UserRagSnapshotService：

  • 专门处理用户级别的快照创建和管理
  • createUserSnapshot() - 为用户创建专属快照
  • copySnapshotToUser() - 从版本快照复制到用户快照

4. 数据访问优化

RagDataAccessService (/AgentX/src/main/java/org/xhy/domain/rag/service/RagDataAccessService.java)

需要完善的方法：

// 当前返回空列表，需要实现
private List<FileDetailEntity> getSnapshotFiles(String versionId)
private List<DocumentUnitEntity> getSnapshotDocuments(String versionId)
private List<DocumentUnitEntity> getSnapshotDocumentsByOriginalFile(String versionId, String originalFileId)

Read the full file on GitHub · 204 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 204 lines · 0 tokens per session scan A 017bb1e29057

Subscribe to this mod's changes

RAG快照机制完善需求文档 is a command published in the GitHub repository lucky-aeon/AgentX (830 stars, last pushed 2mo ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 2,099 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Data and AIRAG and vectors
Related

Other commands, from other repositories

notebooklm

Vault-first source-grounded research via Gemini File Search. One command, no browser. The grounded parallel to /research-deep (which is open-web via Perplexity).

eugeniughelbur/obsidian-second-brain · 35 tokens

bucket-vector

Performs vector similarity search on a bucket using a vector index backed by JVector, with optional post-filtering to combine similarity ranking with structured query predicates.

kronotop/kronotop · 31 tokens

create-kb

Create a KB domain — light single-pass by default, --validated for a source-verified high-assurance build, --audit for KB health.

luanmorenommaciel/agentspec · 32 tokens

ai-pipeline

RAG/embedding pipeline scaffolding — delegates to ai-data-engineer agent.

luanmorenommaciel/agentspec · 19 tokens

agent-brain-index

Index documents for semantic search.

SpillwaveSolutions/agent-brain · 10 tokens

agent-brain-semantic

Search using semantic vector similarity for conceptual queries.

SpillwaveSolutions/agent-brain · 14 tokens