cluster-plannerlisted

SERP analyst and content planner for competitor comparison clusters ({{PRODUCT}} vs {Competitor}). Use this skill when the orchestrator delegates planning, OR when the user asks to "plan a cluster", "research SERPs for", "do keyword research for", "analyze competitors for", "build a content plan for", or wants a content brief for a competitor comparison topic. This skill performs deep SERP analysis on parallel competitor pairs (since head-to-head {{PRODUCT}} vs X queries usually have no organic data), pulls People Also Ask, identifies content gaps, references the product's help center or knowledge base, and outputs a complete cluster plan as both markdown and a Google Sheets-compatible TSV tracker. It reads product-context.md for every product fact and runs a short intake to create it if missing. Trigger this skill aggressively for ANYTHING involving SERP research, content planning, keyword analysis, or hub-and-spoke architecture for competitor comparison clusters. The output is the input to cluster-writer, s
rohitabrahamgeorge/seo-content-team · ★ 0 · Data & Documents · score 72

Install: claude install-skill rohitabrahamgeorge/seo-content-team

# Cluster Planner ## Overview This skill is the analyst layer of the cluster system. It takes a competitor name (e.g. "Globex") and produces a complete content plan that the writer skill can execute against. The planning method is **parallel-competitor SERP mining**: since "{{PRODUCT}} vs Globex" has no organic search history, the planner analyzes how Globex shows up in OTHER comparisons (Globex vs another competitor from the competitor set, and so on) to extract the SERP playbook those competitors use, then maps it onto the PostHog Mixpanel template structure for {{PRODUCT}} to adopt. ## Inputs the orchestrator will provide - `competitor_name` (e.g. "Globex") - `cluster_slug` (e.g. "{{PRODUCT_SLUG}}-vs-globex") - `vertical_context` (optional, e.g. a target industry or buyer segment from the ICP in product-context.md) - `persistence_path` (where to write outputs) If invoked directly without an orchestrator, ask the user for these inputs upfront in a single multi-question prompt. ## Phase 0: Load product context Before anything else, read `product-context.md` at the root of the working folder (next to `clusters/`). It is the single source of truth for the product name, first-mention rule, website, help center URL, positioning, ICP, the three differentiators, verified proof points, verified capabilities by product area, pricing, what the product does NOT have, migration help actually offered, voice, and the competitor set. If `product-context.md` is missing, copy the b