← 返回 Skills 市场
Sentinel QA — World-Class AI Audit & Quality System
作者
tenlifejosh
· GitHub ↗
· v1.0.1
· MIT-0
168
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install sentinel-qa
功能描述
World-class autonomous quality assurance skill system. Use ANY time the user asks to review, check, audit, validate, QA, proofread, inspect, approve, reject,...
安全使用建议
This skill appears to be a thorough QA checklist system and is internally consistent as documentation, but it expects the agent to test and interact with live services (purchases, downloads, email deliverability, analytics) without declaring any credentials or endpoints. Before installing or enabling this skill: 1) Confirm what external integrations your agent already has access to (Gumroad, KDP, email provider, analytics) and who would pay for test purchases. 2) If you want the skill to run checks, provide safe, least-privilege/test/staging credentials (read-only APIs where possible, sandbox payment accounts) and never give production payment credentials without explicit human approval. 3) Require a human confirmation step for any action that would charge money, publish content, or modify live listings. 4) Clarify escalation channels (who/where is 'Hutch') and logging/persistence policies so sensitive data isn't written to unexpected storage. 5) If the skill will be used across organizations, ask the author for a manifest of required integrations and a justification for any account-level access. If you cannot confirm these, treat the skill as potentially able to perform impactful actions and restrict its permissions or run it in a locked/staging environment.
功能分析
Type: OpenClaw Skill
Name: sentinel-qa
Version: 1.0.1
The sentinel-qa skill bundle is a comprehensive quality assurance system designed for an AI agent to review digital products, marketing copy, and brand consistency. The bundle consists entirely of markdown-based checklists and procedural instructions (e.g., in SKILL.md and launch-gates.md) aimed at enforcing high editorial and legal standards for a specific business entity. While product-qa.md contains Python snippets for file hashing and PDF metadata auditing, these are standard utility functions directly related to the stated purpose of verifying file integrity and lack any malicious characteristics such as obfuscation, unauthorized network access, or credential theft.
能力评估
Purpose & Capability
The skill claims to be an all-purpose QA/audit system for published assets (Gumroad products, email campaigns, KDP, social posts, etc.). The SKILL.md and references explicitly require testing live flows (test purchases, verify downloads, check conversion rates, verify SPF/DKIM and sender authentication, run audit sweeps). Yet the skill declares no required environment variables, no config paths, and no binaries. That is inconsistent: platform access and transaction testing normally require credentials and platform-specific access. The mismatch could be benign (the agent environment provides those integrations), but the skill should have declared any needed credentials or scope.
Instruction Scope
Instructions are detailed and checklist-driven (read reference files, run domain-specific checklists, produce PASS/FAIL with evidence). However they also instruct actions that reach beyond local files: 'test purchase to verify flow', 'test email received by test address', 'check sales this month vs last month', 'test purchase completed successfully', and escalate for legal/privacy issues. Those steps imply interacting with external services and possibly performing financial transactions. The skill also refers to persistent failure logs (memory/sentinel-failures.json) and escalation to 'Hutch' without a clear, secure communication mechanism. This breadth of action is not narrow or purely analytical — it can have real-world side effects.
Install Mechanism
Instruction-only skill with no install spec and no code files: lowest installation risk. Nothing in the package attempts to download or execute external code. The references are local documents that the agent is told to read.
Credentials
The skill requests no environment variables or credentials, yet many required checks (platform listing health, purchase flow testing, analytics metrics, SPF/DKIM verification, testing email deliverability) inherently need access to external accounts or APIs. Either the agent will rely on pre-provisioned integrations (not declared here), or the skill will fail or attempt to act with whatever access the agent already has. The lack of declared, proportionate credentials is a mismatch that could lead to unexpected use of other agent-held secrets.
Persistence & Privilege
The skill is not marked always:true and does not install persistent code, but its references describe writing to a failure log and maintaining a memory file (memory/sentinel-failures.json) and creating new QA rules over time. Persisting QA logs is reasonable for a QA skill, but combined with autonomous invocation and the potential to perform actions on live platforms, that persistence increases operational footprint. There is no explicit instruction for requiring human confirmation before executing destructive or billable actions (e.g., purchases).
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 sentinel-qa - 安装完成后，直接呼叫该 Skill 的名称或使用
/sentinel-qa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
v1.0.1
Updated display name.
v1.0.0
World-class quality assurance skill: pre-publish review, product QA, content QA, brand consistency, completeness checks, logic review, risk assessment, launch gates, audit sweeps, failure logging
元数据
常见问题
Sentinel QA — World-Class AI Audit & Quality System 是什么？
World-class autonomous quality assurance skill system. Use ANY time the user asks to review, check, audit, validate, QA, proofread, inspect, approve, reject,... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 168 次。
如何安装 Sentinel QA — World-Class AI Audit & Quality System？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install sentinel-qa」即可一键安装，无需额外配置。
Sentinel QA — World-Class AI Audit & Quality System 是免费的吗？
是的，Sentinel QA — World-Class AI Audit & Quality System 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Sentinel QA — World-Class AI Audit & Quality System 支持哪些平台？
Sentinel QA — World-Class AI Audit & Quality System 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 Sentinel QA — World-Class AI Audit & Quality System？
由 tenlifejosh（@tenlifejosh）开发并维护，当前版本 v1.0.1。
推荐 Skills