360 Secure SkillHub
360 安全技能中心

LLM Signal GEO Analyst

该 Skill 通过 Bash 脚本调用外部 LLM Signal agent 接口，先获取站点 GEO 行动方案、再查询站点状态；仅自动执行 auto_safe 动作，manual/assist 动作转人工审批，并按固定字段输出动作标题、优先级、原因、步骤、命令脚手架与是否需要审批。

immythz
immythz
热度 81.31
最近更新 2026-05-11 17:29:17
分类：营销依赖：第三方 API / 代码执行应用场景：市场分析 / 报告撰写 / 工作流 / API 集成

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add llm-signal-geo-analyst
Metadata
name
llm-signal-geo-analyst
description
Run LLM Signal GEO analyst workflows from OpenClaw. Use when you need to fetch deterministic GEO action plans, check site health status, and execute/review hosted or self-hosted agent workflows with approval-safe guidance.
homepage
https://www.llmsignal.app/docs/agents
user-invocable
true
metadata
{"openclaw":{"homepage":"https://www.llmsignal.app/docs/agents","primaryEnv":"LLMSIGNAL_API_KEY","requires":{"env":["LLMSIGNAL_BASE_URL","LLMSIGNAL_SITE_ID","LLMSIGNAL_API_KEY"],"anyBins":["curl","bash","sh"]}}}

LLM Signal GEO Analyst

Use this skill to operate LLM Signal agent flows from OpenClaw.

Required environment

  • LLMSIGNAL_BASE_URL (example: https://www.llmsignal.app)
  • LLMSIGNAL_SITE_ID
  • LLMSIGNAL_API_KEY

Execution policy

  1. Call POST /api/agent/v1/plan before recommending actions.
  2. Automatically execute only auto_safe actions.
  3. Route manual and assist actions to human approval.
  4. Never output API keys or secrets.

Run plan request

{baseDir}/scripts/fetch-plan.sh

Run status request

{baseDir}/scripts/fetch-status.sh

Output format for each recommended action

Return:

  1. title
  2. priority
  3. reason
  4. exact steps
  5. command/diff scaffold (if present)
  6. approval required (yes for manual/assist, no for auto_safe)

Notes

  • Agent API access is limited to Growth and Pro plans.
  • Use persist=true in plan calls to store run history and outcomes.