GEO generative engine optimization agent
wangzhuo-coding/geo-content-optimizer · tools
GEO generative engine optimization agent — 7 types of keywords + seven-layer architecture + EE-A-T authoritative framework + 8-dimensional mark reduction rewriting
★ 6GitHub stars
2Forks
2026-08-15Last updated
PythonLanguage
MITLicense
Key features
- Multi-format document analysissupports PDF/Word/Excel/PPT/TXT, extracting plain text with one click
- Writing (Stage 3/4) adjudicates user requests according to three levels of priority: Compliance/Factual/Black Hat Red Line (Tier 0) Reject and give compliance substitution
- PIPELINE_QUALITY_JUDGE=falseSkip LLM scoring (only run deterministic lint, saving one LLM round trip)
- 7 types of keyword miningcore search terms, long-tail scene words, regional entity words, crowd label words, problem description words, brand-specific words, scene trigger words, structured output mandatory (Pydantic schema), sorted by priority formula
- Seven-layer architecture content generation: EE-A-T authoritative framework + ACES conclusion pre-paradigm + three-stage Hook opening + semantic entity density network
Install command
dsh plugin --profile web add github:wangzhuo-coding/geo-content-optimizer