← 返回 Skills 市场
User Growth Coach
作者
Jack-Yang-ai
· GitHub ↗
· v2.3.1
· MIT-0
311
总下载
0
收藏
1
当前安装
3
版本数
在 OpenClaw 中安装
/install user-growth-coach
功能描述
三层反馈复盘系统 + 随手记，连接当前输入、历史复盘和日常上下文，自动识别深层行为模式。
安全使用建议
This skill is coherent with its purpose, but it will read your local OpenClaw session transcripts (by default ~/.openclaw/agents/main/sessions) and write structured files under your workspace (default ~/.openclaw/workspace/memory/...). Before installing/run: 1) confirm the session and digest directory paths in SKILL.md/README and, if desired, set OPENCLAW_SESSIONS_DIR / OPENCLAW_DIGEST_DIR to safer locations; 2) back up any existing markdown memory files — migrate-md-to-jsonl.py renames the source file to .bak; 3) review the scripts (they are local and not networked) and test them on a copy of your data to ensure they behave as you expect; 4) if you do not want the skill to process all session transcripts (which may contain sensitive chats), restrict the sessions directory or run extraction scripts manually rather than via an automated cron.
功能分析
Type: OpenClaw Skill
Name: user-growth-coach
Version: 2.3.1
The 'User Growth Coach' skill bundle is a productivity tool designed to analyze user behavior and provide feedback based on session history and manual entries. While the scripts 'extract-raw-inputs.py' and 'extract-daily-digest.py' access sensitive session transcripts located in the OpenClaw sessions directory, this behavior is explicitly documented and necessary for the 'L3' contextual feedback feature. The analysis found no evidence of data exfiltration, unauthorized network access, obfuscation, or malicious command execution; all data processing remains local to the user's environment.
能力评估
Purpose & Capability
The skill claims to implement a three-layer capture/review system and includes scripts and runtime instructions to: capture quick notes, read past JSONL/markdown records, and generate a daily digest from agent session transcripts. The files and scripts present (extract-raw-inputs.py, extract-daily-digest.py, migrate-md-to-jsonl.py, trigger-router.js) match that purpose.
Instruction Scope
SKILL.md instructs the agent to read session transcripts and local memory files (memory/user-growth/* and memory/daily-digest/*) and to run the included scripts to generate digests. Reading transcripts is necessary for L3/day-digest functionality, but it means the skill will process potentially sensitive conversation contents from the agent's sessions directory.
Install Mechanism
No install spec is provided (instruction-only with shipped scripts). That is lower risk: nothing is fetched from the network or installed automatically. The code is included in the bundle and will run locally if invoked.
Credentials
The registry metadata lists no required env vars, but the README/SKILL.md reference optional env vars (OPENCLAW_SESSIONS_DIR and OPENCLAW_DIGEST_DIR) to point at session and digest directories. No secrets or external API keys are requested. The environment access requested is proportional to the stated functionality, though the metadata omission is a minor inconsistency.
Persistence & Privilege
The scripts create and write files under the user's workspace (memory/daily-digest/*.md, memory/user-growth/*.jsonl). migrate-md-to-jsonl.py renames (moves) the input markdown to a .bak backup — i.e., it modifies user files. This file-write and rename behavior is expected for storing digests and migrations but is a material side effect the user should be aware of before running.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install user-growth-coach - 安装完成后，直接呼叫该 Skill 的名称或使用
/user-growth-coach触发 - 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
v2.3.1
v2.3.1 修复元数据字段问题
v2.3.0
Version 2.3.0
- 新增「随手记（Capture）」模式：支持用户白天随时捕获灵感、决定、观察等，零负担，无需深度分析。
- 增加多种「随手记」触发词及自动分类打标签，并极简确认归档。
- 复盘时自动整合当天随手记，输出「今日随手记回顾」并作为L2分析素材。
- 存储结构优化，新增 captures-YYYY-MM.jsonl 文件用于随手记记录。
- 汇总与复盘时可统计随手记分布及与复盘的关联，追踪决策后续执行情况。
- 原有三层（L1/L2/L3）复盘协议及日常摘要流程保持不变。
v2.2.0
User Growth Coach v2.2 introduces a powerful three-layer feedback review system for identifying deep behavioral patterns.
- 全新三层反馈架构：L1（当前输入反馈）、L2（历史行为模式洞察）、L3（日常上下文深度关联）
- 支持多种模式（快速/标准/深度），灵活触发与反馈粒度
- 引入自动行为模式洞察，输出动机解析与具体行动处方
- 日常交互摘要（daily-digest）自动提取与结构化总结，深度补充复盘数据
- 新增成长/复盘汇总输出：跨天模式追踪、承诺清单、建议与明日关注点
- 精细化数据与存储结构，支持高质量推理模型路由
元数据
常见问题
User Growth Coach 是什么？
三层反馈复盘系统 + 随手记，连接当前输入、历史复盘和日常上下文，自动识别深层行为模式。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 311 次。
如何安装 User Growth Coach？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install user-growth-coach」即可一键安装，无需额外配置。
User Growth Coach 是免费的吗？
是的，User Growth Coach 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
User Growth Coach 支持哪些平台？
User Growth Coach 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 User Growth Coach？
由 Jack-Yang-ai（@jack-yang-ai）开发并维护，当前版本 v2.3.1。
推荐 Skills