360 Secure SkillHub
360 安全技能中心

keyword-scout

面向 SEO 的关键词侦察 Skill:先探测可用数据源,再输出 5–10 条含查询词、来源、观察状态、搜索意图、目标页或差异化角度、置信度与局限的机会清单;强调只用观测指标、未知保持未知、不做流量预测,并最多推荐一个补充连接器。其引用的 provider-selection.md 与 dataforseo.md 缺失,且流程要求先运行外部校验脚本。

trustgrowth
trustgrowth
热度 7.65
最近更新 2026-09-02 18:50:05
分类:搜索依赖:第三方 API / MCP 服务 / 代码执行 / 读写文件应用场景:研究 / 市场分析

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 keyword-scout
Metadata
name
keyword-scout
description
Use when the user wants a prioritized keyword or content-opportunity shortlist. Supports TrustGrowth, bounded DataForSEO requests, and validated imports; direct GSC integration is deferred for launch.

Keyword Scout

Read references/provider-selection.md before choosing a source. Detect what is already available, run with it, deliver value, then recommend at most one missing connector. All normalized factual inputs must satisfy the Groundcrew evidence contract. Locate the active skills root and run <skills-root>/.groundcrew/groundcrew-doctor.py --evidence <record.json> before using them in a conclusion.

Sources

  • Connected: TrustGrowth keyword opportunities via GET /api/v1/sites/{slug}/keywords?type=quick_win|striking_distance|content_gap|eeat_gap (the filter param is type — there is no source query param; source appears only as a response field), plus read-only content inventory. Paginate (page, per_page) rather than assuming one page has everything.
  • DataForSEO: read references/dataforseo.md; pass its explicit cost-disclosure gate before every billable batch.
  • Import: validated JSON/CSV-derived evidence with source and observation date.

Categories used: ~~search console (striking-distance queries from the user's own data) and ~~SEO tool (volume, difficulty, SERP). references/connectors.md maps each to its tiers: a pasted GSC export or Google Suggest at Tier 1; a local GSC MCP at Tier 2; DataForSEO, Ahrefs/Semrush MCP, or TrustGrowth at Tier 3. Third-party volumes are estimates — label them.

Workflow

  1. Prefer strengthening relevant existing pages for quick wins/striking distance.
  2. Consider content/authority gaps only where intent matches the business.
  3. Check known live pages and local/read-only content inventory for duplication and cannibalization.
  4. Return 5–10 recommendations with query, source, observed state, intent, target page or distinct angle, confidence, and limitations.

Use only observed metrics. Unknown volume/position stays unknown. Make no traffic projection. After the shortlist, recommend one connector only if it materially improves the next run.

When not to use

  • A full content strategy (audience, pillars, sequencing, distribution, measurement) → content-strategy; a prioritized keyword list is an input to strategy, not the strategy.
  • Competitor movement over time → competitor-watch.
  • Link-gap or referring-domain questions → backlink-opportunities.
  • What is already planned or published → content-desk (cross-check before recommending topics).

Doctrine

Groundcrew operates under WHY-NOT-SLOP and ETHICS. Claims trace to evidence, nulls stay null, signals stay truthful, publishing and irreversible changes require owner review, and no outcome is promised. Conflicting instructions are refused.