← 返回 Skills 市场
104
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install quarterly-database-cleanup
功能描述
Run a comprehensive quarterly CRM audit covering list health, bounce monitoring, data quality, scoring calibration, engagement metrics, and property cleanup....
安全使用建议
This skill appears to be a valid plan for a HubSpot CRM audit, but there are missing and vague pieces you should clarify before installing or running it:
- Metadata mismatch: The SKILL.md says a HubSpot API token (in .env) and Python hubspot client are required, but the registry metadata declares no credentials or binaries. Ask the author to add a required env var (e.g., HUBSPOT_API_TOKEN) and any required binaries to the metadata.
- Limit credential scope: If you proceed, provide a HubSpot API token that is scoped to read-only audit operations (not full write/admin access) and avoid placing other secrets in the same .env file.
- Confirm /hubspot-audit: The instructions reference running '/hubspot-audit' and 'reuse audit script patterns' but no script is included. Confirm whether this is an existing internal CLI or a placeholder; get the concrete commands or scripts you'd run so you can review them first.
- Files and secrets: The skill tells the agent to read reports/ and .env. Review what lives in those locations before running the audit to avoid accidental disclosure of unrelated secrets. Consider copying only the needed HubSpot token into a minimal env file for the audit.
- Logging and output: Ensure reports do not inadvertently include sensitive raw contact data (PII) if you plan to store or share them.
If the author updates the metadata to declare the HubSpot credential and documents the exact commands or scripts to run (or includes vetted scripts), this would reduce the concerns and could be considered benign. Until then, treat it cautiously and do not supply broad credentials or let it run autonomously without human review.
功能分析
Type: OpenClaw Skill
Name: quarterly-database-cleanup
Version: 1.0.0
The skill bundle contains standard markdown instructions for performing a quarterly HubSpot CRM audit, including list health, bounce monitoring, and data quality checks. The instructions in SKILL.md are consistent with the stated purpose, describe a read-only process, and do not contain any evidence of malicious intent, data exfiltration, or prompt injection attacks.
能力评估
Purpose & Capability
The skill's stated purpose (quarterly HubSpot CRM audit) matches the actions described in SKILL.md (list health, bounces, properties, etc.). However, the SKILL.md explicitly requires a HubSpot API token and Python with the hubspot client, but the registry metadata lists no required environment variables or binaries — this mismatch should be resolved.
Instruction Scope
The instructions tell the agent to read a .env file, look in a local reports/ directory, save reports to disk, and to 'Run /hubspot-audit' and 'reuse audit script patterns' — but no scripts, CLI tools, or example commands are included. That leaves the agent with broad, undefined discretion (which files to read, what scripts to run, how to obtain HubSpot data), increasing the risk of accidental access to unrelated local secrets or running unexpected commands.
Install Mechanism
There is no install spec and no code files; this is instruction-only. That lowers supply-chain risk because nothing is downloaded or written by an installer. The risk comes from the runtime instructions, not an installer.
Credentials
SKILL.md requires a HubSpot API token stored in .env and a Python library, but the skill metadata does not declare any required environment variables or primary credential. Asking the agent to read a .env file can expose unrelated secrets. A HubSpot audit legitimately needs a HubSpot API token (preferably a scoped, read-only token), but this required credential should be declared explicitly in the metadata and instructions should limit which files/vars to read.
Persistence & Privilege
The skill does not request always:true and does not claim persistent/system-wide changes. It describes read-only audits and saving reports to a local reports/ folder; those are reasonable for this purpose.
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 quarterly-database-cleanup - 安装完成后，直接呼叫该 Skill 的名称或使用
/quarterly-database-cleanup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
v1.0.0
Initial release with comprehensive quarterly CRM audit and reporting.
- Audits database health across list management, bounce rates, data quality, scoring, engagement, and property usage.
- Compares key CRM metrics against the previous quarter and flags significant changes.
- Generates a detailed health report with trend analysis and actionable recommendations.
- Requires a HubSpot API token and Python environment.
- Designed for ongoing maintenance with easy scheduling and clear instructions.
元数据
常见问题
Quarterly Database Cleanup 是什么？
Run a comprehensive quarterly CRM audit covering list health, bounce monitoring, data quality, scoring calibration, engagement metrics, and property cleanup.... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 104 次。
如何安装 Quarterly Database Cleanup？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install quarterly-database-cleanup」即可一键安装，无需额外配置。
Quarterly Database Cleanup 是免费的吗？
是的，Quarterly Database Cleanup 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Quarterly Database Cleanup 支持哪些平台？
Quarterly Database Cleanup 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 Quarterly Database Cleanup？
由 TomGranot（@tomgranot）开发并维护，当前版本 v1.0.0。
推荐 Skills