by 安全扫描
OpenClaw
安全
high confidenceThe skill is an instruction-only complaint-handling playbook centered on Linear; its requirements and instructions are internally consistent and do not request unexplained credentials or external installs.
评估建议
This skill is a textual playbook for handling member complaints via Linear and appears coherent. Before installing or enabling it for autonomous use, confirm: (1) how your platform supplies Linear access (connectors/credentials) so you don’t accidentally allow write permissions you didn’t intend, (2) whether you want the agent to post comments automatically or only prepare drafts for human review, and (3) that any company SOPs or sensitive customer data referenced are provided to the skill or re...详细分析 ▾
ℹ 用途与能力
The skill's name and description (Linear-centered complaint handling) match the SKILL.md content. It expects to read from and write back to Linear issues. The skill declares no required env vars or binaries — this is plausible for an instruction-only skill that assumes platform connectors will provide Linear access, but the omission is worth noting if you expect the skill to autonomously call Linear's API.
✓ 指令范围
The runtime instructions stay within the stated purpose: parse Linear issue metadata, produce an analysis comment and a customer-reply draft, and generate digests. There are no instructions to read arbitrary host files, exfiltrate data, call unknown endpoints, or invent policies. The only external action implied is reading/writing Linear issues, which is consistent with the described operating model.
✓ 安装机制
No install spec or code files are included; the skill is instruction-only, so nothing will be downloaded or written to disk by the skill itself.
ℹ 凭证需求
The skill requests no environment variables or credentials. This is proportionate for an instruction-only playbook, but it implicitly assumes the agent or platform will provide Linear integration/credentials if automated read/write is required — confirm how Linear access is granted on your platform before using it for live writes.
✓ 持久化与权限
always:false and default model-invocation behavior are both set; the skill does not request forced/global inclusion or elevated persistence. It does imply potential autonomous action (normal platform default), so confirm runtime policies for automatic posting to Linear if that is a concern.
安全有层次，运行前请审查代码。
运行时依赖
无特殊依赖
版本
latestv1.0.02026/3/17
- Initial release of member-complaint-agent skill. - Manages customer complaint workflows centered on Linear, with structured analysis and reply drafting. - Parses deterministic metadata (e.g., platform, vendor, app version) separately from AI-based judgment. - Provides two default Linear comments per complaint: structured AI analysis and a ready-to-use customer reply draft. - Includes detailed guidelines for SOP handling, escalation, and daily/weekly digest summaries. - Enforces strict separation of facts, judgments, and recommended actions; prohibits unauthorized promises.
● 无害
安装命令
点击复制官方npx clawhub@latest install member-complaint-agent
镜像加速npx clawhub@latest install member-complaint-agent --registry https://cn.longxiaskill.com