← Back to Skills Marketplace
team-memory
by
jichengkai
· GitHub ↗
· v2.3.0
· MIT-0
213
Downloads
1
Stars
0
Active Installs
5
Versions
Install in OpenClaw
/install team-memory
Description
私人团队记忆管理系统 - 记录成员观察、生成绩效报告、辅助管理决策
Usage Guidance
This skill is internally consistent for building a local team memory system: it stores and reads markdown files and its scripts only create directories/files under ~/.config/opencode/skills/team-memory. Before installing or running scripts: 1) Inspect scripts/init.sh and scripts/new-member.sh yourself (they just mkdir and write files) and run them in a non-privileged account. 2) Fix ambiguous aliases in skill-config.yaml (e.g., "ZS" appears for multiple members) to avoid misattribution. 3) Treat the data as sensitive: do not enable cloud-sync unless you intentionally want backups in a cloud service and you understand encryption/retention/consent/legal obligations for personnel data. 4) Because the package source/homepage is unknown, prefer running locally without network access and keep backups under your control; if you need long-term support, prefer a skill with a known maintainer or host the files in an internal repo. 5) If you intend to share or archive records, adopt a retention/deletion policy and document consent for recorded observations.
Capability Analysis
Type: OpenClaw Skill
Name: team-memory
Version: 2.3.0
The skill bundle is a team management system designed to store sensitive PII and performance observations locally. While the intent appears legitimate, the shell scripts (specifically scripts/new-member.sh) are vulnerable to path traversal and potential command injection because they do not sanitize input variables like 'NAME' when creating files or directories. Additionally, the installation instructions in README.md and scripts/init.sh rely on hardcoded paths (e.g., ~/Desktop), and the system encourages the creation of 'shadow files' containing sensitive HR-like data without any built-in encryption or access controls.
Capability Assessment
Purpose & Capability
Name/description (team memory for managers) match the included files and scripts: markdown data files, templates, an init script, and a new-member script. The skill does not request unrelated credentials or binaries. One operational caveat: the package source/homepage is unknown which affects trust/maintenance but does not make the capability incoherent.
Instruction Scope
SKILL.md explicitly describes reading and using local markdown files (蒸馏/时间轴/档案) which is consistent with the stated purpose. It also documents optional workflows that involve copying/syncing data to cloud storage (README examples show Dropbox/网盘 symlink) — those are user-driven actions and not automatic, but they materially affect privacy. Additionally, skill-config.yaml / shortcuts contain a potential alias conflict (alias "ZS" used for multiple members), which could cause misattribution when using shortcuts; this is a configuration quality issue you should fix before use.
Install Mechanism
No install spec in the registry; the skill is instruction + two small shell scripts. The scripts only create directories and write markdown files under $HOME/.config/opencode/skills/team-memory and do not download or execute remote code. No high-risk install behavior detected.
Credentials
The skill declares no required environment variables, no primary credential, and no config paths outside its own ~/.config/opencode/skills/team-memory area. The example config has a backup.cloud-sync flag defaulting to false; enabling cloud sync would be a user decision and increases exposure risk but is not required by the skill.
Persistence & Privilege
The skill does not request 'always: true' and does not modify other skills. Its scripts persist data under the user's home config directory (expected for this purpose). Autonomous invocation is allowed (platform default) but there is no evidence the skill abuses persistence or escalates privileges.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install team-memory - After installation, invoke the skill by name or use
/team-memory - Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.3.0
- 新增“向上管理”与“公司动向”模块,支持记录上级期望与公司战略变化
- 新增数据目录 data/upward/上级期望.md 和 data/company/战略方向.md
- SKILL.md 说明文档同步更新,介绍新模块和数据结构
- 支持更完整的团队上下管理视角
v2.2.3
- 引入 Structured Distillation 三层架构:档案层、蒸馏层、原始层,提升成员管理的结构化程度
- 新增成员档案(含性格、OKR、发展计划)和蒸馏摘要文件,便于快速理解和检索
- 数据结构优化,成员内容分为“档案.md”“蒸馏.md”“时间轴.md”三个独立文件
- 移除旧有时间轴格式使用指南,添加双层架构场景示例
- 使用节奏与模板结构更加聚焦 OKR、年度沟通和多层记录
v2.2.1
Version 2.2.1
- 新增参考文档:场景示例、记录模板、故障排除(references/*.md)
- 增加脚本工具:目录初始化脚本与新成员创建脚本(scripts/init.sh, scripts/new-member.sh)
- 增强数据结构,新增“管理反思”和自动追踪目录
- 简化并明确三种记录类型(观察、对话、管理)
- 更新使用流程和隐私说明,提升易用性
- 移除过多细节,提升文档简洁度
v2.2.0
✅ SKILL.md 从 408行 → 90行(精简 78%)
✅ 示例和模板移到 references/
✅ 创建初始化脚本
✅ README.md 更新结构说明
v2.1.0
team-memory v2.1.0
- 全新重写SKILL.md,详细说明本地团队记忆系统的功能、数据结构和典型场景
- 明确聚焦于为技术管理者提供日常观察、结构化绩效分析及多场景报告输出
- 补充了事件结构、记录流程、快速记录模板和输出模板
- 加入隐私与数据本地化保护建议
- 新增团队健康度仪表盘与预测性提醒等进阶管理辅助玩法
- 用户指南覆盖初始化、日常流程、常见问题与改进建议
Metadata
Frequently Asked Questions
What is team-memory?
私人团队记忆管理系统 - 记录成员观察、生成绩效报告、辅助管理决策. It is an AI Agent Skill for Claude Code / OpenClaw, with 213 downloads so far.
How do I install team-memory?
Run "/install team-memory" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is team-memory free?
Yes, team-memory is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does team-memory support?
team-memory is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created team-memory?
It is built and maintained by jichengkai (@jichengkai); the current version is v2.3.0.
More Skills