← 返回 Skills 市场
Self-Improving Operations
作者
José I. O.
· GitHub ↗
· v1.1.0
· MIT-0
101
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install self-improving-operations
功能描述
Captures process bottlenecks, incident patterns, capacity issues, automation gaps, SLA breaches, and toil accumulation to enable continuous operations improv...
安全使用建议
This skill appears coherent and low-risk for its stated goal of capturing and promoting operations learnings. Before enabling it: (1) review the scripts (scripts/*.sh) and hook handlers to ensure you’re comfortable with their behavior; (2) prefer enabling the activator (UserPromptSubmit) only rather than PostToolUse if your tool outputs may contain sensitive data, since error-detector reads CLAUDE_TOOL_OUTPUT; (3) install hooks only into intended user or project-level hook directories (avoid blindly copying into global ~/.openclaw/hooks unless intended); (4) keep the promise in SKILL.md: do not record secrets, credentials, internal IPs, or customer PII into .learnings/ files. If you need higher assurance, run the scripts locally in a sandboxed repo first and confirm they only write the expected files.
功能分析
Type: OpenClaw Skill
Name: self-improving-operations
Version: 1.1.0
The 'self-improving-operations' skill bundle is designed to help AI agents log and analyze operational bottlenecks, incidents, and toil. It includes utility scripts (activator.sh, error-detector.sh) for monitoring command outputs and a scaffolding script (extract-skill.sh) that features basic path sanitization to prevent directory traversal. The instructions in SKILL.md and the OpenClaw hooks (handler.js/ts) focus on structured logging and explicitly command the agent to redact secrets, credentials, and PII, showing a clear alignment with its stated purpose without malicious intent.
能力评估
Purpose & Capability
The name/description (capture and promote operational learnings) matches the provided assets, templates, runbook helpers, and hook code. The included scripts (activator, error detector, extractor) and hook handlers are consistent with a workflow that reminds agents to log learnings, scans tool output for operational error patterns, and scaffolds new skills from learnings.
Instruction Scope
Runtime instructions and hooks operate on .learnings/ files in the project or OpenClaw workspace and inject reminder content at agent bootstrap; they do not instruct reading unrelated system files or external endpoints. Note: the error-detector hook reads the CLAUDE_TOOL_OUTPUT environment variable (expected in PostToolUse hook context) — this can contain tool output and may include sensitive snippets if not redacted. The SKILL.md explicitly warns not to log secrets/PII, which is appropriate.
Install Mechanism
There is no formal install spec (instruction-only), which is low risk. However, the package contains executable scripts and hook handlers; enabling the hooks or copying files into ~/.openclaw/ (or other agent hooks directories) will persist those scripts into the user's environment. The manual install suggestion uses a public GitHub URL; no opaque download or extract-from-untrusted-URL behavior is present.
Credentials
The skill requires no credentials, binaries, or config paths. It references CLAUDE_TOOL_OUTPUT (hook-provided context) and writes/creates .learnings/ files — both are proportional to its stated purpose. No secret or unrelated credentials are requested.
Persistence & Privilege
always is false and the skill is user-invocable; hooks are opt-in. The hook handler injects a virtual reminder file at bootstrap and scripts operate within the workspace or relative paths; the skill does not modify other skills' configs or request system-wide privileges.
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 self-improving-operations - 安装完成后，直接呼叫该 Skill 的名称或使用
/self-improving-operations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
v1.1.0
**Version 1.1.0**
- Added stackability contract for multi-skill installations.
- Added namespaced logging guidance (`.learnings/operations/`) for coexistence with other skills.
- Added required `Skill: operations` metadata field and cross-skill precedence/ownership rules.
- Clarified hook arbitration model (single dispatcher, dedupe, rate limiting).
v1.0.0
- Initial release of the self-improving-operations skill.
- Provides structured workflow and templates for capturing operational learnings, incident patterns, capacity issues, automation gaps, SLA breaches, and toil accumulation.
- Guides logging and escalation of repeat incidents, MTTR breaches, manual steps, alert fatigue, change failure spikes, and other key operational signals.
- Includes setup instructions for OpenClaw and generic agent integration.
- Promotes persistent knowledge sharing through categorized markdown log files, with guidelines for promoting significant learnings to runbooks, postmortems, and backlogs.
- Offers optional session-start reminders via hook integration.
元数据
常见问题
Self-Improving Operations 是什么？
Captures process bottlenecks, incident patterns, capacity issues, automation gaps, SLA breaches, and toil accumulation to enable continuous operations improv... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 101 次。
如何安装 Self-Improving Operations？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install self-improving-operations」即可一键安装，无需额外配置。
Self-Improving Operations 是免费的吗？
是的，Self-Improving Operations 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Self-Improving Operations 支持哪些平台？
Self-Improving Operations 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 Self-Improving Operations？
由 José I. O.（@jose-compu）开发并维护，当前版本 v1.1.0。
推荐 Skills