← 返回 Skills 市场
Self-Improving Analytics
作者
José I. O.
· GitHub ↗
· v1.1.0
· MIT-0
94
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install self-improving-analytics
功能描述
Captures data quality issues, metric drift, pipeline failures, misleading visualizations, metric definition mismatches, and data freshness problems to enable...
安全使用建议
This skill appears coherent and does what it says: it scaffolds .learnings/ logs and injects lightweight reminders via optional hooks. Before enabling: (1) review the scripts (scripts/activator.sh, scripts/error-detector.sh, scripts/extract-skill.sh) to confirm you’re comfortable with their behavior; (2) prefer enabling the UserPromptSubmit activator only (not PostToolUse) unless you want automatic error detection from tool output — PostToolUse reads CLAUDE_TOOL_OUTPUT and may run for many commands; (3) enable hooks per-project or add matcher filters so reminders only run for analytics-related sessions; (4) check file permissions and the paths where the hook is copied (~/.openclaw/hooks, ~/.openclaw/workspace) to avoid global changes you didn’t intend; (5) never copy unreviewed SQL or outputs containing credentials/PII into .learnings/ — the SKILL.md explicitly warns about this. If you want extra assurance, run the extract-skill.sh with --dry-run and test the activator locally before enabling hooks globally.
功能分析
Type: OpenClaw Skill
Name: self-improving-analytics
Version: 1.1.0
The self-improving-analytics skill bundle is designed to help AI agents track and resolve data quality issues, metric drift, and pipeline failures within an analytics workspace. It utilizes shell scripts (activator.sh, error-detector.sh, extract-skill.sh) and OpenClaw hooks (handler.js) to provide automated reminders and scaffold new skills based on captured learnings. The bundle follows security best practices by explicitly instructing the agent not to log credentials or PII, and the scripts include basic path sanitization and input validation to prevent common command injection risks.
能力评估
Purpose & Capability
The name/description match the provided artifacts: markdown templates, reminder scripts, a hook handler, and helpers to scaffold/promote analytics learnings. There are no unrelated required env vars, binaries, or external download/install steps that would be inconsistent with a logging/reminder skill.
Instruction Scope
SKILL.md and scripts focus on creating/maintaining .learnings/ files and injecting reminders via OpenClaw hooks. The activator and error-detector are intended to be run as hooks and the error detector reads the CLAUDE_TOOL_OUTPUT environment variable to detect errors; SKILL.md also instructs copying hooks into ~/.openclaw/hooks and enabling them. These actions affect the agent workspace and hook configuration (expected for this skill) — review before enabling, and be cautious when enabling PostToolUse hooks since they run on tool output.
Install Mechanism
There is no install spec that downloads or executes remote code. The package is instruction-plus-local-scripts and hook handlers that run locally. All files are present in the skill bundle; nothing pulls from arbitrary URLs or writes unexpected binaries.
Credentials
The skill declares no required environment variables or credentials. Scripts do read CLAUDE_TOOL_OUTPUT (an agent-provided variable) to detect errors — this is reasonable for a PostToolUse detector but is not listed as a required env var. The README cautions not to log credentials or PII. Examples reference OpenClaw functions (sessions_send/sessions_spawn) which are platform features, not additional secrets.
Persistence & Privilege
always:false and hooks are opt-in. The skill suggests copying hook files into ~/.openclaw/hooks and enabling them, which is expected for a hook; it does not demand forced always-on installation or modification of other skills' configuration beyond registering its own hook.
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-analytics - 安装完成后，直接呼叫该 Skill 的名称或使用
/self-improving-analytics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/analytics/`) for coexistence with other skills.
- Added required `Skill: analytics` 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-analytics skill for structured logging and continuous analytics improvement.
- Enables standardized logging of data quality issues, metric drift, pipeline failures, schema changes, and misleading visualizations to Markdown files.
- Provides initialization instructions for a `.learnings/` directory with pre-defined files: `LEARNINGS.md`, `DATA_ISSUES.md`, and `FEATURE_REQUESTS.md`.
- Outlines clear use cases and quick-reference guidance for when and where to log specific analytics incidents.
- Supports promotion of important learnings to data dictionaries, pipeline runbooks, dashboard standards, or SLAs for persistent improvement.
- Includes OpenClaw-specific setup and generic integration instructions for other agents.
- Provides detailed logging format and workflow to ensure actionable, privacy-conscious analytics learnings.
元数据
常见问题
Self-Improving Analytics 是什么？
Captures data quality issues, metric drift, pipeline failures, misleading visualizations, metric definition mismatches, and data freshness problems to enable... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 94 次。
如何安装 Self-Improving Analytics？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install self-improving-analytics」即可一键安装，无需额外配置。
Self-Improving Analytics 是免费的吗？
是的，Self-Improving Analytics 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Self-Improving Analytics 支持哪些平台？
Self-Improving Analytics 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 Analytics？
由 José I. O.（@jose-compu）开发并维护，当前版本 v1.1.0。
推荐 Skills