← 返回 Skills 市场
stephongao

six-dimension-evolution

作者 stephonGAO · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
117
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install six-dimension-evolution
功能描述
Six-dimension evolution system for AI agents. Transform from reactive assistant to proactive partner with lessons tracking, success patterns, decision review...
安全使用建议
This skill is a template-based 'agent self-improvement' system and will copy multiple files into your ~/.openclaw/workspace, write an installed timestamp, and add/modify HEARTBEAT.md to schedule daily heartbeats. Before installing: 1) Back up your workspace and HEARTBEAT.md; 2) Review the template files to confirm you’re comfortable with automated daily reads of memory/session files and automatic writes to SKILL_STATS.md, MEMORY.md, etc.; 3) Ask the skill author how report pushing is implemented — the SKILL.md mentions sending full reports via Feishu but provides no instructions or required credentials (this is an incoherence and a potential data‑exfiltration vector if misconfigured); 4) If you don’t want automatic scheduling or external pushes, decline activation or ask for a version that only provides templates without modifying HEARTBEAT.md or pushing reports; 5) If you install, monitor the first activation closely and verify what is sent externally and to which endpoints.
功能分析
Type: OpenClaw Skill Name: six-dimension-evolution Version: 1.0.0 The 'six-dimension-evolution' skill bundle is a comprehensive framework designed to help AI agents track lessons, successes, user preferences, and knowledge gaps. It uses SKILL.md to instruct the agent to perform a setup process that includes copying template files to the workspace and modifying the HEARTBEAT.md file for scheduled task execution. While the system performs significant file management and session analysis (via 1-session-summary.md and 2-daily-review.md), these actions are transparently documented, aligned with the stated purpose of 'self-evolution,' and include explicit instructions for the agent to seek user confirmation before activation.
能力评估
Purpose & Capability
The name/description (agent self‑improvement) aligns with the actions the skill prescribes: creating LESSONS/SUCCESSES/DECISIONS/PREFERENCES/SKILL_STATS/KNOWLEDGE_GAPS and heartbeat files and then performing daily reviews. Creating and updating workspace files and logs is proportionate to the stated purpose. One mismatch: the SKILL.md explicitly instructs pushing complete reports via Feishu (飞书) but the skill declares no environment variables or credentials for an external push service.
Instruction Scope
The instructions mandate enumerating and reading session files, memory files, and many workspace/config files and then writing/patching HEARTBEAT.md and SKILL_STATS.md (installed timestamp). Those actions are within a self‑improvement system, but the skill also requires producing and '推送完整报告' (push full reports) externally without specifying how to authenticate or which endpoint — this creates a risk of leaking sensitive conversation data if the agent is configured to forward reports. The instructions also insist on modifying HEARTBEAT.md scheduling rules automatically, which changes agent behavior and scheduling without further checks beyond the initial prompt.
Install Mechanism
Instruction-only skill with no install spec and no code to download or execute; lowest install risk. It performs file operations (cp/mkdir) from the local templates directory to the user's workspace, which is expected for a template-based skill.
Credentials
The skill references external push (飞书 Interactive Card) and scheduling triggers but declares zero required environment variables and no credentials. If the skill expects to push to Feishu or other external systems, it should declare required tokens/URLs. Also it expects write access to multiple workspace/config files (HEARTBEAT.md, SKILL_STATS.md, config/heartbeat-state.json, memory/*.md), which is reasonable functionally but is a privilege the user should explicitly consent to.
Persistence & Privilege
The skill does not set always:true and is user-invocable only. However it instructs writing an 'installed' timestamp and modifying HEARTBEAT.md to add scheduled heartbeat tasks — that grants it persistent presence via workspace files and scheduled behavior. This is explainable for an evolution system but the user should be aware it modifies scheduling/config so future autonomous actions (daily heartbeats) may be triggered by those files.
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 six-dimension-evolution
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /six-dimension-evolution 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
v1.0.0
Initial release: introduces a comprehensive six-dimension evolution system for AI agents. - Enables agents to transform from reactive assistants to proactive partners with six evolution files: lessons tracking, success patterns, decision review, preference tracking, skill statistics, and knowledge gap detection. - Provides an activation and setup workflow with workspace file checks, user prompts, and command instructions. - Integrates with daily and session review scheduling via HEARTBEAT.md for continuous self-improvement. - Includes detailed documentation for use cases, file structure, and customization options.
元数据
Slug six-dimension-evolution
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

six-dimension-evolution 是什么?

Six-dimension evolution system for AI agents. Transform from reactive assistant to proactive partner with lessons tracking, success patterns, decision review... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 117 次。

如何安装 six-dimension-evolution?

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

six-dimension-evolution 是免费的吗?

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

six-dimension-evolution 支持哪些平台?

six-dimension-evolution 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 six-dimension-evolution?

由 stephonGAO(@stephongao)开发并维护,当前版本 v1.0.0。

💬 留言讨论