← 返回 Skills 市场
swaylq

Session Memory

作者 Sway Liu · GitHub ↗ · v2.0.0 · MIT-0
cross-platform ⚠ suspicious
5573
总下载
1
收藏
53
当前安装
3
版本数
在 OpenClaw 中安装
/install session-memory
功能描述
Persistent memory toolkit for AI agents. Save context, recall with relevance scoring, consolidate insights, track decisions across sessions. Features importa...
安全使用建议
Install only if you are comfortable keeping agent memories as local plaintext files. Do not store API keys, passwords, tokens, or raw sensitive details in this memory; use vault references instead. Avoid letting untrusted prompts, imported backups, or external content choose script arguments until the input-interpolation bugs are fixed, and treat exports/backups as sensitive files.
功能分析
Type: OpenClaw Skill Name: session-memory Version: 2.0.0 The skill bundle provides a functional local memory system but contains multiple security vulnerabilities due to a lack of input sanitization. Specifically, 'daily.sh' is vulnerable to path traversal via the date argument, and 'edit.sh' and 'recall.sh' are susceptible to shell/grep injection because user-provided variables are passed directly into shell commands. While the intent appears to be a legitimate utility for agent continuity, these flaws allow for unauthorized file access or command execution if the agent is prompted with crafted inputs.
能力评估
Purpose & Capability
The stated purpose is persistent local agent memory, and the bundled commands match that purpose: save, recall, context loading, consolidation, stats, export/import, pruning, and edit/delete.
Instruction Scope
Several scripts interpolate command-line values directly into inline Node.js source instead of passing them only as argv data, so crafted values for options such as limits, days, topics, or dates can escape their intended scope and run local JavaScript under the user account.
Install Mechanism
The package declares only a Node binary requirement, has no npm dependencies or install lifecycle hooks, and does not show background installation behavior.
Credentials
Writing to ~/.agent-memory is proportionate for a memory skill, but the storage is plaintext JSONL and the docs do not clearly warn against storing secrets or sensitive operational details.
Persistence & Privilege
Persistence is disclosed and central to the skill, but high and critical memories are automatically resurfaced into future context, and export/prune/delete features lack strong warnings or safeguards against sensitive-data exposure or accidental loss.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install session-memory
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /session-memory 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v2.0.0
v2.0.0: relevance search, importance levels, context loader, consolidation, export/import, stats, edit/delete
v1.1.0
Fix: safe JSON via node (no more manual sed escaping), fix glob depth for recall/topics, add --json mode to all commands, add missing prune.sh, fix requires declaration
v1.0.0
Initial release: Persistent memory toolkit for AI agents
元数据
Slug session-memory
版本 2.0.0
许可证 MIT-0
累计安装 191
当前安装数 53
历史版本数 3
常见问题

Session Memory 是什么?

Persistent memory toolkit for AI agents. Save context, recall with relevance scoring, consolidate insights, track decisions across sessions. Features importa... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 5573 次。

如何安装 Session Memory?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install session-memory」即可一键安装,无需额外配置。

Session Memory 是免费的吗?

是的,Session Memory 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Session Memory 支持哪些平台?

Session Memory 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Session Memory?

由 Sway Liu(@swaylq)开发并维护,当前版本 v2.0.0。

💬 留言讨论