agentleFS
matteotittamatteotitta/genesys-skills/skills/aeo-strategy/SKILL.md

aeo-strategy

Skill53 starsChanged 5 days agoMITView on GitHub

Plans content that gets a company cited by ChatGPT, Perplexity and Google AI: topic clusters, keyword gaps, competitor content analysis, funnel-stage allocation and a dated article queue. Use when asked for an AEO strategy, AI search content plan or what to publish to get cited.

The file

Files
SKILL.md

AEO Strategy

Produce a research-backed content strategy for search + AI visibility. Output: cluster taxonomy, keyword gap analysis, competitor content audit, prioritised article queue with target keywords, and a 90-day publishing timeline. It decides what to write; writing the articles is a separate step.

When to run

User says "AEO strategy", "content roadmap", "keyword gap analysis", "what should [company] publish", "90-day content plan for search", "article queue for [company]", or "what content will get us cited by AI?". Skip if they want a single article, a content plan across every channel, or a pure SEO audit. Full trigger list: the premium reference.

Inputs

Required: company context (product, features) · 3-5 competitors with URLs · GSC exports (Queries CSV, Pages CSV, Coverage) · GA4 exports (Traffic acquisition, Landing pages). Without GSC/GA4, all keyword data must be marked [ESTIMATED] and strategy confidence flagged in the executive summary.

Recommended: ICP notes · an existing content audit · positioning and messaging · competitor notes · a content plan · sales call transcripts (these enable Phase 2.5 query seeding).

Optional: prior keyword research · monthly content target (override default 20) · AirOps brand kit.

Full export instructions (exact GSC/GA4 click paths, columns, used-in-phase mapping) + validation checklist: the premium reference.

Steps

Eight phases. Each phase has sub-steps, output templates, and checkpoints in the premium reference — read it before producing the output.

  1. Phase 1 — Input validation. Load company context · confirm 3-5 competitors with geography tags · set parameters (default: 20/month, 90 days, 50/20/30 BOFU/MOFU/TOFU split; adjust by maturity).
  2. Phase 2 — Cluster taxonomy. Build 5-8 product-mapped pillars (not generic categories), each with 5-10 sub-clusters from product docs + ICP pains + competitor content.
  3. Phase 2.5 — Transcript query seeding (only if you have sales call transcripts). Mine buyer questions, convert to natural-language queries, tag by cluster + source, merge into aeo/query-index.md (or seed it).
  4. Phase 3 — SEO keyword gap analysis. Pull client baseline (GSC) + competitor keywords (DataForSEO → Apify → Exa fallback) · identify gaps · group by cluster · separate by geography · pull AirOps AI-citation baseline · write 1-paragraph competitor keyword strategy summaries.
  5. Phase 5 — Competitor best performing content. Exa + Firecrawl + DataForSEO traffic estimates → pick top 5 per competitor · tag each with funnel stage + cluster · write 1-paragraph strategy analysis per competitor.
  6. Phase 6 — Content type strategy. Define types per stage (BOFU: comparisons, branded, integrations, pricing, demo · MOFU: how-to, "Best X for Y", deep-dives, use cases, compliance · TOFU: definitions, industry guides, regulatory, thought leadership) · set % allocation grounded in Phase 4 gaps · write rationale per type.
  7. Phase 7 — Article queue. Generate specific titles per type with target keywords. BOFU: 1 comparison per competitor + branded pages. MOFU: how-tos mapped to features + "Best for [year]" listicles, flag UPDATE vs CREATE. TOFU: definitions targeting highest-volume gap keywords + industry/regulatory guides.
  8. Phase 8 — 90-day timeline. Assign articles to months (M1: comparisons + high-volume TOFU + initial MOFU · M2: BOFU deep-dives + continue MOFU/TOFU · M3: remaining BOFU + integrations + TOFU depth) · build period × stage summary table.

Anti-hallucination guardrails (must apply during execution): keyword volumes sourced or marked [ESTIMATED: based on SERP analysis] · page counts from crawl data only · no invented competitor content (verify via Exa/Firecrawl) · article titles flagged as suggestions · explicitly note approximated research. Full guardrails + per-phase checkpoint lists: the premium reference and the premium reference.

What good looks like

Output structure (the premium reference for full template): # [X]-day content roadmap for [Company] → Executive summary (2-3 sentences + content-type table) → Research (A. clusters · B. keyword gaps · C. competitor content · D. competitor best performing) → Strategy and content examples (rationale + article tables) → Timeline (monthly + summary) → Iteration prompts.

Design principles: executive summary table first (10-second scan) · research before strategy · clean tables over prose walls · specific ready-to-brief titles with target keywords (not categories) · simple monthly timeline.

Pre-delivery quality gates (full checklists in the premium reference):

  • Research: 5-8 clusters mapped to product features · keyword gaps with volumes · competitor crawl page counts · pages classified by stage AND cluster · top 5 per competitor.
  • Strategy: allocation sums to 100% · monthly counts realistic · rationale grounded in Phases 3-5 (not generic) · types appropriate for client maturity.
  • Execution: specific titles (not categories) · target keywords per article · UPDATE vs CREATE flags · monthly priorities actionable immediately · summary totals mathematically correct.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.