Ragtop Agent

高级 RAG 助手，具备 Agentic RAG 思考能力。能够自动化管理 RAGTOP 知识库，并在执行深度调研任务时，调用 RAGTOP 后端接口。使用场景包括：(1) 列出知识库和文档，(2) 执行语义检索和深度分析。

This skill is largely coherent with its stated purpose, but take these precautions before installing or enabling it: - Verify the RAGTOP_API_URL you intend the skill to call; the SKILL.md defaults to an internal IP (http://10.71.10.71:9380). If that is not your RAGTOP service, set RAGTOP_API_URL explicitly — otherwise the agent will make network calls to that address. - Limit the RAGTOP_API_TOKEN privileges (least privilege) and use a token scoped only to the KBs/doc access needed. Prefer short‑lived tokens and rotation. - Ask the publisher for source/homepage or provenance; the registry metadata is missing RAGTOP_API_URL and there is no homepage, which reduces transparency. - Monitor network traffic and logs after enabling, and consider running the agent in a restricted environment until you confirm behavior. If you cannot confirm the default endpoint or token scope, treat this skill as risky and do not supply sensitive credentials.

Type: OpenClaw Skill Name: ragtop-agent Version: 1.0.0 The skill instructs the AI agent to execute `curl` commands directly in a shell, interpolating environment variables like `RAGTOP_API_URL` and `RAGTOP_API_TOKEN` (found in `SKILL.md` and `references/workflow.md`). This pattern introduces a shell injection vulnerability if the OpenClaw platform does not adequately sanitize or escape these environment variables before execution. Additionally, the default `RAGTOP_API_URL` points to a private IP address (`http://10.71.10.71:9380`), indicating internal network access capabilities, which, while plausible for its stated purpose, represents a risky capability.