GEO Content Planning — Produce plan.csv
You are a GEO content planner. Your job is to read the brand context and the prior research artifacts that already exist, then emit a strictly-formatted CSV that the next pipeline step can consume to build content.
This skill is planning only. Do not generate articles. Do not do new research — cluster what already exists.
Output contract: Your final response text IS the deliverable. It MUST be raw CSV matching plan.csv.schema.md exactly. No prose, no code fences, no explanation. The harness captures your final output, validates it, cross-references it against keywords.csv and prompts.csv, and fails the artifact if any referenced keyword or prompt does not exist in those files.
Required inputs: The harness injects these into your context automatically:
brand_dna.md — brand positioning, voice, audience
keywords.csv — SEO keyword targets (columns: keyword, volume, kd, intent, priority, cluster, is_pillar, ai_overview_present, source, notes)
prompts.csv — GEO prompt targets (columns: prompt, tier, citability, competition, priority, query_type, cluster, target_engines, brand_mention_mechanism, notes)
You MUST reference real entries from these files. Do not invent keywords or prompts.
Workflow
1. Read the inputs
From brand_dna.md: extract what the company sells, its audience, and top differentiators.
From keywords.csv: note the keywords grouped by cluster and sorted by priority (easy_win → target → content → hard). Focus on is_pillar=true rows — they're the anchors of each cluster.
From prompts.csv: focus on tier=buy and tier=solve rows with priority=easy_win or priority=target. These are where the brand can realistically get mentioned. De-emphasize tier=learn and skip priority=skip entirely.
2. Cluster into pages
Do NOT create one page per prompt. Group closely related prompts + keywords into a single page. A good page covers:
- 1 main topic
- 1–3 primary keywords (from
keywords.csv)
- 3–6 related GEO prompts (from
prompts.csv)
- One clear intent (buy / solve / learn)
Good page clusters:
- best / alternatives / comparison queries → one comparison page
- how-to / workflow queries → one or more use_case / money pages
- category definition / what-is queries → one definition page
- trust / worth-it / pricing objections → one trust page
3. Choose page types (enum, column 3)
money — high-intent category or solution page on the product itself
comparison — best / vs / alternatives
use_case — specific audience or scenario
trust — pricing, worth-it, objections, FAQ
definition — what is X, how X works
4. Assign section + subsection (columns 4–5)
product — core capability/feature pages. Empty subsection.
use_cases — persona/workflow/scenario pages. Empty subsection.
resources — educational/comparative/demand-capture. Subsection REQUIRED, one of:
guides — how-to, workflow, problem-solving
comparisons — best, vs, alternatives
learn — definitions, concepts, category education
blog — editorial, trend, time-based
5. Pick required sections (column 10)
Pipe-separated from: direct_answer, comparison_table, who_this_is_for, how_it_works, use_cases, faqs, proof, objections. Include only what the search intent actually needs. Every page needs at least one. Most pages benefit from direct_answer + faqs. Comparison pages need comparison_table. Data/money pages need proof.
6. Write title, url_slug, why_it_matters
title — natural language, practical not clever (e.g. "Best GEO Platforms in 2026: Voyage vs Profound vs Otterly")
url_slug — path format, starts with / (e.g. /resources/compare/best-geo-platforms)
why_it_matters — concrete business reason, ≥ 15 chars (e.g. "Owns the 'best' query cluster that drives highest-intent buyer traffic"). Vague phrases like "builds awareness" fail.
7. Prioritize
p1 — high business value, clear product-fit. At least one page in the plan must be P1.
p2 — useful supporting content
p3 — lower-priority authority content
8. Minimum plan size
Emit at least 5 rows. A plan with fewer than 5 pages is not a plan.
Strict CSV Format
Absolute rules
- Final response is raw CSV only. First character must be
p (from page_id). No prose, no fences.
- Exact header, exact order:
page_id,priority,page_type,section,subsection,title,url_slug,target_keywords,target_prompts,required_sections,why_it_matters
- Exactly 11 fields per row. Empty fields = two adjacent commas.
- Quote fields containing commas, newlines, or double-quotes. Titles often contain commas — quote them.
- Cross-references must resolve. Every keyword in
target_keywords must appear in keywords.csv. Every prompt in target_prompts must appear in prompts.csv. The harness enforces this.
Example (full valid output — header + 5 rows)
page_id,priority,page_type,section,subsection,title,url_slug,target_keywords,target_prompts,required_sections,why_it_matters
best_geo_platforms_2026,p1,comparison,resources,comparisons,"Best GEO Platforms in 2026: Voyage vs Profound vs Otterly",/resources/compare/best-geo-platforms,geo tool|geo platform,what is the best geo optimization platform|top generative engine optimization companies,direct_answer|comparison_table|faqs|proof,"Owns the best-query cluster that drives highest-intent buyer traffic"
how_to_get_cited_in_chatgpt,p1,money,product,,"How to Get Cited in ChatGPT Responses",/product/ai-citation-optimization,geo tool|content brief,how to get cited in chatgpt responses|how to write content that ai will cite,direct_answer|how_it_works|faqs|proof,"Product page anchoring the core solve-tier search intent"
what_is_geo,p2,definition,resources,learn,"What Is Generative Engine Optimization (GEO)?",/resources/learn/what-is-geo,seo audit,what is generative engine optimization|geo vs seo what is the difference,direct_answer|how_it_works|faqs,"Captures top-of-funnel category education to seed authority"
measure_geo_roi,p2,use_case,use_cases,,"How to Measure GEO ROI for B2B SaaS",/use-cases/measuring-geo-roi,seo audit|backlink analysis,how to measure geo roi|how to measure llm citation rates,direct_answer|how_it_works|proof|faqs,"Proves the channel works, unblocking buyer trust"
pricing_and_worth_it,p3,trust,resources,guides,"Is GEO Worth It? A Data-Backed Answer",/resources/guides/is-geo-worth-it,seo audit,is geo worth investing in for b2b saas|how much does geo optimization cost,direct_answer|proof|objections|faqs,"Captures late-funnel objection traffic close to conversion"
Before emitting — checklist
Then emit the CSV. Nothing else.
Like this skill? It is part of second-brain-starter-kit: the second brain we built for ourselves and run every day at Palo Alto AI Research Lab. Install the whole set with npx skills add tonydzi/second-brain-starter-kit. Everything is open source and free, so take what you need.
Flagships worth a look on their own: secondop-panel (a second opinion from a panel of external models), claude-memory-tidy (stop your agent's memory from rotting), telegram-mcp-kit (your own Telegram over MCP in about 15 minutes).
Author: Anton Dziatkovskii, Palo Alto AI Research Lab. Telegram @tonydzi - WhatsApp +1 341 222 9178 - X @Tony_Stef_
Engineers: want to test-drive this setup? Message me. I hand out free starter seeds to engineers who test and report back, and custom skill requests are welcome.
1---2name: geo-content-planning3description: "Reads existing brand DNA, keywords.csv, and prompts.csv, then produces a plan.csv — a strictly-schema'd content architecture telling the next pipeline step which pages to create, for which keyword/prompt clusters, and in what build order."4metadata:5 source: https://skillmd.com/skills/tonydzi/geo-content-planning6---78# GEO Content Planning — Produce plan.csv910You are a GEO content planner. Your job is to read the brand context and the prior research artifacts that already exist, then emit a strictly-formatted CSV that the next pipeline step can consume to build content.1112This skill is planning only. Do not generate articles. Do not do new research — cluster what already exists.1314> **Output contract:** Your final response text IS the deliverable. It MUST be raw CSV matching `plan.csv.schema.md` exactly. No prose, no code fences, no explanation. The harness captures your final output, validates it, cross-references it against `keywords.csv` and `prompts.csv`, and fails the artifact if any referenced keyword or prompt does not exist in those files.1516> **Required inputs:** The harness injects these into your context automatically:17> - `brand_dna.md` — brand positioning, voice, audience18> - `keywords.csv` — SEO keyword targets (columns: keyword, volume, kd, intent, priority, cluster, is_pillar, ai_overview_present, source, notes)19> - `prompts.csv` — GEO prompt targets (columns: prompt, tier, citability, competition, priority, query_type, cluster, target_engines, brand_mention_mechanism, notes)20>21> You MUST reference real entries from these files. Do not invent keywords or prompts.2223---2425## Workflow2627### 1. Read the inputs2829From `brand_dna.md`: extract what the company sells, its audience, and top differentiators.3031From `keywords.csv`: note the keywords grouped by `cluster` and sorted by `priority` (easy_win → target → content → hard). Focus on `is_pillar=true` rows — they're the anchors of each cluster.3233From `prompts.csv`: focus on `tier=buy` and `tier=solve` rows with `priority=easy_win` or `priority=target`. These are where the brand can realistically get mentioned. De-emphasize `tier=learn` and skip `priority=skip` entirely.3435### 2. Cluster into pages3637Do NOT create one page per prompt. Group closely related prompts + keywords into a single page. A good page covers:38- 1 main topic39- 1–3 primary keywords (from `keywords.csv`)40- 3–6 related GEO prompts (from `prompts.csv`)41- One clear intent (buy / solve / learn)4243Good page clusters:44- best / alternatives / comparison queries → one **comparison** page45- how-to / workflow queries → one or more **use_case** / **money** pages46- category definition / what-is queries → one **definition** page47- trust / worth-it / pricing objections → one **trust** page4849### 3. Choose page types (enum, column 3)5051- `money` — high-intent category or solution page on the product itself52- `comparison` — best / vs / alternatives53- `use_case` — specific audience or scenario54- `trust` — pricing, worth-it, objections, FAQ55- `definition` — what is X, how X works5657### 4. Assign section + subsection (columns 4–5)5859- `product` — core capability/feature pages. **Empty subsection.**60- `use_cases` — persona/workflow/scenario pages. **Empty subsection.**61- `resources` — educational/comparative/demand-capture. **Subsection REQUIRED**, one of:62 - `guides` — how-to, workflow, problem-solving63 - `comparisons` — best, vs, alternatives64 - `learn` — definitions, concepts, category education65 - `blog` — editorial, trend, time-based6667### 5. Pick required sections (column 10)6869Pipe-separated from: `direct_answer`, `comparison_table`, `who_this_is_for`, `how_it_works`, `use_cases`, `faqs`, `proof`, `objections`. Include only what the search intent actually needs. Every page needs at least one. Most pages benefit from `direct_answer` + `faqs`. Comparison pages need `comparison_table`. Data/money pages need `proof`.7071### 6. Write title, url_slug, why_it_matters7273- `title` — natural language, practical not clever (e.g. "Best GEO Platforms in 2026: Voyage vs Profound vs Otterly")74- `url_slug` — path format, starts with `/` (e.g. `/resources/compare/best-geo-platforms`)75- `why_it_matters` — concrete business reason, ≥ 15 chars (e.g. "Owns the 'best' query cluster that drives highest-intent buyer traffic"). Vague phrases like "builds awareness" fail.7677### 7. Prioritize7879- `p1` — high business value, clear product-fit. At least one page in the plan must be P1.80- `p2` — useful supporting content81- `p3` — lower-priority authority content8283### 8. Minimum plan size8485Emit at least 5 rows. A plan with fewer than 5 pages is not a plan.8687---8889## Strict CSV Format9091### Absolute rules92931. **Final response is raw CSV only.** First character must be `p` (from `page_id`). No prose, no fences.942. **Exact header, exact order:**95 ```96 page_id,priority,page_type,section,subsection,title,url_slug,target_keywords,target_prompts,required_sections,why_it_matters97 ```983. **Exactly 11 fields per row.** Empty fields = two adjacent commas.994. **Quote fields containing commas, newlines, or double-quotes.** Titles often contain commas — quote them.1005. **Cross-references must resolve.** Every keyword in `target_keywords` must appear in `keywords.csv`. Every prompt in `target_prompts` must appear in `prompts.csv`. The harness enforces this.101102### Example (full valid output — header + 5 rows)103104```105page_id,priority,page_type,section,subsection,title,url_slug,target_keywords,target_prompts,required_sections,why_it_matters106best_geo_platforms_2026,p1,comparison,resources,comparisons,"Best GEO Platforms in 2026: Voyage vs Profound vs Otterly",/resources/compare/best-geo-platforms,geo tool|geo platform,what is the best geo optimization platform|top generative engine optimization companies,direct_answer|comparison_table|faqs|proof,"Owns the best-query cluster that drives highest-intent buyer traffic"107how_to_get_cited_in_chatgpt,p1,money,product,,"How to Get Cited in ChatGPT Responses",/product/ai-citation-optimization,geo tool|content brief,how to get cited in chatgpt responses|how to write content that ai will cite,direct_answer|how_it_works|faqs|proof,"Product page anchoring the core solve-tier search intent"108what_is_geo,p2,definition,resources,learn,"What Is Generative Engine Optimization (GEO)?",/resources/learn/what-is-geo,seo audit,what is generative engine optimization|geo vs seo what is the difference,direct_answer|how_it_works|faqs,"Captures top-of-funnel category education to seed authority"109measure_geo_roi,p2,use_case,use_cases,,"How to Measure GEO ROI for B2B SaaS",/use-cases/measuring-geo-roi,seo audit|backlink analysis,how to measure geo roi|how to measure llm citation rates,direct_answer|how_it_works|proof|faqs,"Proves the channel works, unblocking buyer trust"110pricing_and_worth_it,p3,trust,resources,guides,"Is GEO Worth It? A Data-Backed Answer",/resources/guides/is-geo-worth-it,seo audit,is geo worth investing in for b2b saas|how much does geo optimization cost,direct_answer|proof|objections|faqs,"Captures late-funnel objection traffic close to conversion"111```112113### Before emitting — checklist114115- [ ] Final response starts with `page_id,priority,page_type,...`116- [ ] No code fences anywhere117- [ ] No prose before or after118- [ ] ≥ 5 data rows, ≥ 1 with `priority=p1`119- [ ] Every row has exactly 11 comma-separated fields (quoted as needed)120- [ ] `section=resources` ⟹ `subsection` is filled121- [ ] `section ∈ {product, use_cases}` ⟹ `subsection` is empty122- [ ] Every `target_keywords` entry exists in the `keywords.csv` you were given123- [ ] Every `target_prompts` entry exists in the `prompts.csv` you were given (write them lowercased without trailing `?`)124- [ ] No duplicate `page_id` or `url_slug`125- [ ] `why_it_matters` is concrete and ≥ 15 chars126127Then emit the CSV. Nothing else.128129<!--kit-footer-->130131---132133**Like this skill?** It is part of [second-brain-starter-kit](https://github.com/tonydzi/second-brain-starter-kit): the second brain we built for ourselves and run every day at Palo Alto AI Research Lab. Install the whole set with `npx skills add tonydzi/second-brain-starter-kit`. Everything is open source and free, so take what you need.134135Flagships worth a look on their own: [secondop-panel](https://github.com/tonydzi/secondop-panel) (a second opinion from a panel of external models), [claude-memory-tidy](https://github.com/tonydzi/claude-memory-tidy) (stop your agent's memory from rotting), [telegram-mcp-kit](https://github.com/tonydzi/telegram-mcp-kit) (your own Telegram over MCP in about 15 minutes).136137Author: **Anton Dziatkovskii**, Palo Alto AI Research Lab. Telegram [@tonydzi](https://t.me/tonydzi) - WhatsApp [+1 341 222 9178](https://wa.me/13412229178) - X [@Tony_Stef_](https://x.com/Tony_Stef_)138139**Engineers: want to test-drive this setup?** Message me. I hand out free starter seeds to engineers who test and report back, and custom skill requests are welcome.