yaojingang

geo-discover

Discover evidence-aware GEO questions, query rewrites, intent clusters, and prioritized content opportunities from a structured brief. Use for AI search intent mining, question or keyword expansion, query research, FAQ discovery, 拓词, and GEO topic discovery in Chinese or English.

yaojingang 136 23 Updated 3w ago

Resources

6
GitHub

Install

npx skillscat add yaojingang/geohub/geo-discover

Install via the SkillsCat registry.

About this skill

We need to produce a 2-3 sentence plain-text summary, objective, factual, no marketing language, no superlatives, no calls to action. At most 60 words. Should explain what the skill does, what problem it solves, when to use it. From description: skill discovers evidence-aware GEO questions, query rewrites, intent clusters, prioritized content opportunities from a structured brief. Used for AI search intent mining, question/keyword expansion, query research, FAQ discovery, 拓词, GEO topic discovery in Chinese or English.

SKILL.md

GEO Discover

Workflow

  1. Read references/discovery-method.md; read references/discovery-method-v2.md when requesting a non-legacy execution mode.
  2. Prepare a protocol 1.0.0 file-backed GEO brief.
  3. Run python3 scripts/run_discover.py --input <brief.json> --output <runs-root> --execution-mode <legacy|deterministic|research|provider>. Legacy remains the compatibility default.
  4. Inspect quality-report.json and run-lineage.json; surface all warnings and failed checks.
  5. Deliver the Artifact Bus directory as the output contract.

Output contract

Treat the output argument as the runs root. Produce <runs-root>/<run-id>/input/geo-brief.json, run-manifest.json, run-lineage.json, evidence-ledger.json, query-map.json, opportunity-map.json, and quality-report.json. Return the actual <runs-root>/<run-id> directory.

Boundaries

Legacy and deterministic modes are offline and file-backed. Research mode consumes only approved evidence already present in the brief. Provider mode needs an explicit runtime adapter; the packaged CLI has no adapter and degrades to deterministic generation with a recorded failure. Discovery records missing evidence and never invents platform responses, volume, ranking, conversion, competitor, or customer data. Diagnosis and content generation remain separate Registry stages.