← 返回 Skills 市场
chameleon-nexus

Geo Opt Coordinator

作者 chameleon-nexus · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ 安全检测通过
37
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install geo-opt-coordinator
功能描述
GEO 智能优化任务 OpenClaw 协同 — 拉取 OPT 周期状态,分发深度仿写与 sau 群发。
使用说明 (SKILL.md)

\r \r

GEO 智能优化协同(Coordinator)\r

\r 轮询 SaaS 已开启 OpenClaw 协同OPT-* 任务,按服务端返回的 openclawActions 执行子流程。\r \r

前置\r

\r

  • ~/.qclaw/geo-api-key(QClaw;Windows %USERPROFILE%\.qclaw\geo-api-key),或兼容路径 ~/.openclaw/geo-api-key(商户 API Key,在 SaaS 账户设置 → 龙虾密钥 自助签发)\r
  • SaaS 创建优化任务时勾选 深度仿写 和/或 sau 平台(会自动 openclawCoordinatorEnabled=true)\r
  • Celery 已跑完当轮诊断(needsDeepImitate 依赖 diagnosis_report_id)\r
  • 深度仿写:本机 FIRECRAWL_API_KEY + firecrawl_scrape 工具\r
  • sau 发布:本机安装 social-auto-upload,完成各平台 Cookie 登录\r \r

启动后自动轮询\r

\r OpenClaw 启动后,建议使用 loop 定时执行本 skill 完整循环(无需 SaaS Webhook):\r \r

/loop 60s 执行 geo-opt-coordinator 推荐循环:拉取 OPT 任务,处理 needsDeepImitate 与 needsSauPublish\r
```\r
\r
或将以下流程写入 Agent 系统提示词 / heartbeat,每 60s 执行一次。\r
\r
## API\r
\r
- 基础地址:`https://ai.gaobobo.cn`(本地 `http://127.0.0.1:8002`)\r
- 鉴权:`Authorization: Bearer $GEO_KEY`\r
\r
### 拉取任务列表\r
\r
```bash\r
GEO_KEY=$(cat ~/.qclaw/geo-api-key 2>/dev/null || cat ~/.openclaw/geo-api-key 2>/dev/null)\r
curl -s "https://ai.gaobobo.cn/api/geo/optimization/tasks" \\r
  -H "Authorization: Bearer $GEO_KEY"\r
```\r
\r
响应每项含:\r
\r
- `task`:品牌、关键词、开关\r
- `latestCycle`:本轮 `contentTaskIds`、`cycleStepResults`\r
- `openclawActions`:\r
  - `needsDeepImitate` → 使用 skill **geo-deep-imitate**\r
  - `needsSauPublish` + `pendingSauPlatforms` → 使用 skill **geo-social-publish**\r
  - `publishableContentTaskIds`:可 sau 分发的 `CG-*`\r
\r
### 单任务详情\r
\r
```bash\r
curl -s "https://ai.gaobobo.cn/api/geo/optimization/OPT-XXXXXXXXXXXX" \\r
  -H "Authorization: Bearer $GEO_KEY"\r
```\r
\r
### 最新周期\r
\r
```bash\r
curl -s "https://ai.gaobobo.cn/api/geo/optimization/OPT-XXXXXXXXXXXX/cycles/latest" \\r
  -H "Authorization: Bearer $GEO_KEY"\r
```\r
\r
## 推荐循环\r
\r
1. `GET /optimization/tasks`\r
2. 对 `needsDeepImitate=true` 的任务执行 **geo-deep-imitate**(同一 `taskId` + `latestCycle.cycleNumber`)\r
3. 对 `needsSauPublish=true` 的任务,对每个 `publishableContentTaskIds` × `pendingSauPlatforms` 执行 **geo-social-publish**\r
4. 间隔 ≥ 60s 再次拉取(文章轮询仍须 ≥30s)\r
\r
## 回写与 SaaS 可见性\r
\r
- 深度仿写完成后 Backend 写入 `latestCycle.cycleStepResults.deepImitate`(status、taskIds、usedSources)\r
- sau 发布回写 `cycleStepResults.sauPublish.platforms.{slug}` 与发稿记录\r
- SaaS 优化任务监控 / 周期日志中展示 **OpenClaw 协同** 区块\r
\r
## 注意\r
\r
- 不替代 SaaS Celery 优化周期;不调用网易 OAuth 发稿\r
- 遗留全局 Key 无法使用协同接口,须商户 Key\r
安全使用建议
Install only if you intend this agent to poll your GEO SaaS tasks and automatically trigger deep imitation and social publishing for enabled OPT tasks. Use a scoped merchant API key where possible, review which SaaS tasks have OpenClaw coordination enabled, and be careful with social-auto-upload cookies because they can authorize public posting on connected platforms.
能力标签
requires-oauth-tokenrequires-sensitive-credentials
能力评估
Purpose & Capability
The skill's purpose is coherent with its capabilities: it polls OPT tasks from a SaaS API, then delegates deep imitation and sau social publishing based on server-returned openclawActions. Social publishing and content generation are high-impact actions, but they are explicitly described and tied to the stated workflow.
Instruction Scope
The runtime instructions recommend a 60-second loop or heartbeat and operate on SaaS tasks with OpenClaw coordination enabled. Scope is server-directed rather than hidden, though users should understand it may repeatedly process all eligible OPT tasks.
Install Mechanism
The artifact contains only a SKILL.md markdown instruction file, with no executable scripts, declared dependencies, or install-time code.
Credentials
It requires a merchant API key, optional Firecrawl API key, and social-auto-upload cookie login for publishing. These credentials are proportionate to the stated integrations and are disclosed in the prerequisites.
Persistence & Privilege
No self-persistence or privilege escalation is present, but the recommended loop creates repeated automated polling and publishing behavior under the user's configured agent environment.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install geo-opt-coordinator
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /geo-opt-coordinator 触发
  4. 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 of geo-opt-coordinator. - Supports polling SaaS for active "OPT-*" tasks with OpenClaw collaboration enabled. - Automates handling of deep imitation and sau publishing actions based on backend guidance. - Provides recommended loop logic for periodic retrieval and processing of optimization tasks. - Documents required authentication and API endpoints for integration. - Outlines expected backend writes and SaaS-side visibility for deep imitation and sau publishing status.
元数据
Slug geo-opt-coordinator
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Geo Opt Coordinator 是什么?

GEO 智能优化任务 OpenClaw 协同 — 拉取 OPT 周期状态,分发深度仿写与 sau 群发。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 37 次。

如何安装 Geo Opt Coordinator?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install geo-opt-coordinator」即可一键安装,无需额外配置。

Geo Opt Coordinator 是免费的吗?

是的,Geo Opt Coordinator 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Geo Opt Coordinator 支持哪些平台?

Geo Opt Coordinator 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 Geo Opt Coordinator?

由 chameleon-nexus(@chameleon-nexus)开发并维护,当前版本 v1.0.0。

💬 留言讨论