Skills agentdb-memory-patterns

📦

agentdb-memory-patterns

Name: agentdb-memory-patterns
Author: DNYoussef

Content revision r1 Medium Risk ⚙️ External commands🌐 Network access

使用 AgentDB 建構代理記憶

具狀態的代理需要在工作階段與工具之間具備可靠的記憶。此技能提供 AgentDB 模式，用於儲存、擷取、遷移與最佳化代理記憶。

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Maintain this Skill? Claim it ⚠️ 38 Poor

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "agentdb-memory-patterns" from https://skillstore.io/skills/dnyoussef-agentdb-memory-patterns.md and its manifest at https://skillstore.io/api/skills/dnyoussef-agentdb-memory-patterns/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/dnyoussef-agentdb-memory-patterns.md Signed manifest GET /api/skills/dnyoussef-agentdb-memory-patterns/manifest Signed lockfile GET /api/skills/dnyoussef-agentdb-memory-patterns/lockfile

Test it

Using "agentdb-memory-patterns". 為支援聊天機器人設計工作階段記憶。

Expected outcome:

一份記憶計畫，包含已儲存的訊息欄位、工作階段篩選器、擷取限制與保留注意事項。

Using "agentdb-memory-patterns". 為程式設計助理規劃長期偏好儲存。

Expected outcome:

一個事實模型，包含分類、信心水準、來源追蹤，以及使用者偏好的更新規則。

Using "agentdb-memory-patterns". 準備 ReasoningBank 到 AgentDB 的遷移審查。

Expected outcome:

一份遷移檢查清單，涵蓋備份、驗證、敏感記憶處理、復原與效能檢查。

Security Audit

Medium Risk

v8 • 7/5/2026 Open versioned report

Most static Ruby backtick detections are Markdown fences, TypeScript snippets, or template literals and are false positives. Confirmed issues remain where the skill instructs users to run unpinned npx commands, register an MCP server, and handle memory database exports or migrations. No prompt injection or malicious exfiltration intent was found.

Files scanned

340

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (6)

Medium

SKILL.md:72

SQLite database file

dbPath: '.agentdb/reasoningbank.db',

The dbPath config targets a persistent ReasoningBank database used for conversation memory. That local store can contain user interactions, so sensitive data handling is real.

Medium

SKILL.md:208

SQLite database file

npx agentdb@latest export ./agents.db ./backup.json

The export command copies the local AgentDB database to backup.json. A memory backup can contain embeddings or conversation-derived data, so mishandling can disclose sensitive contents.

Medium

SKILL.md:236

SQLite database file

'.swarm/memory.db', // Source (legacy)

The migration reads a legacy memory database that can hold prior agent memory. It is legitimate, but it processes sensitive local memory data.

Medium

SKILL.md:237

SQLite database file

'.agentdb/reasoningbank.db' // Destination (AgentDB)

The migration writes the AgentDB destination for persistent memory. It is legitimate, but it stores sensitive agent memory that needs retention protections.

Medium

SKILL.md:323

SQLite database file

npx agentdb@latest migrate --source .swarm/memory.db

The troubleshooting migration command reads .swarm/memory.db through the CLI. Legacy memory can contain sensitive sessions, so the data handling risk is real.

Medium

SKILL.md:79-91 SKILL.md:134-140 SKILL.md:206-211

Persistent Memory Privacy Guidance Missing

The skill stores conversation text and user preference data, then shows export and import workflows. It does not give guidance for consent, retention, encryption, deletion, or secure backup handling.

The cited sections directly store conversation and preference data and show database export/import commands. The file contains no matching privacy or retention guidance.

Capability review items (10)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Medium

SKILL.md:24-36

Ruby/shell backtick execution

```bash

This fenced CLI block instructs users to run npx agentdb@latest init, which downloads and executes an external npm package. Because latest is unpinned, package compromise or version drift could affect users.

Medium

SKILL.md:40-46

Ruby/shell backtick execution

```bash

This CLI block starts and registers an AgentDB MCP server through npx and Claude Code. An unpinned MCP server expands the agent tool surface, so users need explicit trust and version pinning.

Medium

SKILL.md:50-63

Ruby/shell backtick execution

```bash

This CLI block runs npx agentdb@latest create-plugin, which executes an external package and writes project files. The behavior is legitimate but still an external command execution risk.

Medium

SKILL.md:190-202

Ruby/shell backtick execution

```bash

This block runs npx agentdb@latest query against a local memory database. It executes an unpinned package and may expose local memory contents in command output.

Medium

SKILL.md:206-215

Ruby/shell backtick execution

```bash

This block runs npx import, export, and stats commands on local memory and backup files. The export path can copy sensitive memory data, and the package is unpinned.

Medium

SKILL.md:219-227

Ruby/shell backtick execution

```bash

This block runs npx agentdb@latest benchmark, which executes an external npm package. It appears legitimate but still carries supply-chain risk from latest.

Medium

SKILL.md:276-285

Ruby/shell backtick execution

```bash

This block runs plugin management commands through unpinned npx. Even read-only plugin inspection executes external package code.

Medium

SKILL.md:306-312

Ruby/shell backtick execution

```bash

This troubleshooting block runs an unpinned npx stats command against a local database. It is benign in intent but still external command execution.

Medium

SKILL.md:321-324

Ruby/shell backtick execution

```bash

This block runs an unpinned migration command against a legacy memory database. It executes package code while reading persistent memory data.

Medium

SKILL.md:324-338

Ruby/shell backtick execution

```

This range includes an inline npx agentdb@latest mcp command in the Learn More section. It is documentation, but it still encourages starting an unpinned external MCP server.

Audited by: codex View Audit History →

Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report

Copy report link

https://skillstore.io/skills/dnyoussef-agentdb-memory-patterns/audits/8?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/dnyoussef-agentdb-memory-patterns/security.svg)](https://skillstore.io/skills/dnyoussef-agentdb-memory-patterns?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/dnyoussef-agentdb-memory-patterns?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/dnyoussef-agentdb-memory-patterns/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/dnyoussef-agentdb-memory-patterns.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

Academic citations (APA · BibTeX · CFF)

APA citation

DNYoussef. (2026). agentdb-memory-patterns security audit report (audit version 8) [Author version unspecified]. Skillstore. https://skillstore.io/skills/dnyoussef-agentdb-memory-patterns/audits/8

BibTeX citation

@techreport{dnyoussef-dnyoussef-agentdb-memory-patterns-2026,
  author = {DNYoussef},
  title = {agentdb-memory-patterns security audit report (audit version 8)},
  institution = {Skillstore},
  year = {2026},
  number = {8},
  url = {https://skillstore.io/skills/dnyoussef-agentdb-memory-patterns/audits/8},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "agentdb-memory-patterns security audit report (audit version 8)"
version: "unspecified"
type: report
authors:
  - name: "DNYoussef"
date-released: "2026-07-05"
url: "https://skillstore.io/skills/dnyoussef-agentdb-memory-patterns/audits/8"
identifiers:
  - type: other
    value: "skillstore:dnyoussef-agentdb-memory-patterns:audit:8"
    description: "Skillstore immutable audit report identifier"

Compare variants

2 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Highest Skillstore Score

Author / source Skillstore Score Evidence Confidence Skillstore usage Updated

ruvnet Recommended

ruvnet-agentdb-memory-patterns

Skillstore Score 75

Evidence Confidence High

Skillstore usage 14

Updated

2026-09-09

DNYoussef Current

dnyoussef-agentdb-memory-patterns

Skillstore Score 38

Evidence Confidence High

Skillstore usage 14

Updated

2026-09-09

Skillstore Score

Why this score Evidence Confidence: High

Architecture

Maintainability

Content

Community

Spec Compliance

What You Can Build

新增工作階段記憶

為具狀態的聊天代理設計訊息歷史儲存與擷取。

設定 Claude Code 記憶工具

連接 AgentDB MCP server，讓 Claude Code 可以使用持久記憶工作流程。

遷移 ReasoningBank 資料

規劃從舊版 ReasoningBank 儲存遷移到 AgentDB 支援的記憶。

Try These Prompts

規劃基本工作階段記憶

使用此技能為聊天代理設計工作階段記憶。包含儲存欄位、擷取限制與失敗處理。

設計長期事實

使用此技能為我的助理定義長期使用者事實。包含分類、信心規則與更新行為。

建構模式學習工作流程

使用此技能建立模式學習工作流程。包含觸發條件、成功指標、整合時機與擷取策略。

審查記憶遷移計畫

使用此技能審查我的 ReasoningBank 遷移計畫。識別資料風險、驗證步驟、復原步驟與效能檢查。

Best Practices

在執行 CLI 或 MCP 指令前，釘選 AgentDB 套件版本。
將記憶資料庫與匯出檔視為敏感使用者資料。
在儲存對話前，定義保留、刪除與同意規則。

Avoid

不要在未經使用者核准的情況下自動執行 latest 套件指令。
不要將記憶備份匯出到共享或公開資料夾。
不要在缺乏明確目的與保留規則的情況下儲存私人使用者事實。

Frequently Asked Questions

此技能有助於建構什麼？

它協助使用 AgentDB、本機資料庫與 ReasoningBank 模式，為 AI 代理設計持久記憶。

此技能是否包含可執行的應用程式程式碼？

否。它提供文件、CLI 範例與 TypeScript 模式，供使用者調整並套用到自己的專案。

它可以與 Claude Code 搭配使用嗎？

可以。此技能記錄了用於 Claude Code 整合的 AgentDB MCP server 工作流程。

AgentDB memory 可以包含哪些資料？

它可以包含對話歷史、使用者偏好、學習到的模式、embeddings，以及遷移後的代理記憶。

使用前，使用者應驗證什麼？

使用者應驗證套件版本、MCP 權限、資料保留政策與備份儲存控制。

它適合用於正式環境記憶系統嗎？

它可以引導正式環境設計，但團隊必須加入隱私、安全性、監控與營運控制。

Developer Details

Author

DNYoussef

License

MIT

Skillstore revision

Version notice

The author did not declare a version.

Repository

https://github.com/DNYoussef/ai-chrome-extension/tree/main/.claude/skills/agentdb-memory-patterns

Ref

c579efd02184f1128d6741d16e6c8d951d7d4208

Maintenance freshness

7/21/2026

Usage

7 downloads · 266 views

File structure

📄 SKILL.md

agentdb-memory-patterns

Install with my Agent

Agent-readable resources

Test it

Security Audit

Confirmed security concerns (6)

Risk Factors

Copy report link

Markdown badge

HTML badge

Embed card

APA citation

BibTeX citation

CITATION.cff

Compare variants

Skillstore Score

What You Can Build

新增工作階段記憶

設定 Claude Code 記憶工具

遷移 ReasoningBank 資料

Try These Prompts

Best Practices

Avoid

Frequently Asked Questions

Developer Details

More from DNYoussef

slash-command-encoder

micro-skill-creator

cascade-orchestrator

web-cli-teleport

agentdb-semantic-vector-search

when-creating-presentations-use-pptx-generation

building-agents

agent-development

building-commands

analyzing-component-quality

improving-components

skill-builder

agentdb-memory-patterns

Install with my Agent

Agent-readable resources

Test it

Security Audit

Confirmed security concerns (6)

Risk Factors

Copy report link

Markdown badge

HTML badge

Embed card

APA citation

BibTeX citation

CITATION.cff

Compare variants

Skillstore Score

What You Can Build

新增工作階段記憶

設定 Claude Code 記憶工具

遷移 ReasoningBank 資料

Try These Prompts

Best Practices

Avoid

Frequently Asked Questions

Developer Details

More from DNYoussef

slash-command-encoder

micro-skill-creator

cascade-orchestrator

web-cli-teleport

agentdb-semantic-vector-search

when-creating-presentations-use-pptx-generation

Related Skills

building-agents

agent-development

building-commands

analyzing-component-quality

improving-components

skill-builder