LeadUp Blog Content Writer
Purpose
Take a topic or target keyword and produce a complete blog (or service) content package: brief, SEO outline, title options, meta description, FAQ, internal link suggestions, and a final human-style draft ready to publish on a LeadUp client website or SaaS site.
When to use
Use when the user wants a single post or page written. Do not trigger
for keyword research only (use leadup-keyword-competitor-researcher), an
SEO audit (use leadup-seo-strategist), or a content calendar
(use leadup-content-calendar-builder).
Trigger phrases: "write blog", "seo blog", "article writer", "blog content", "write service page", "human blog", "draft this post", "write an article on".
Inputs needed
- Topic or target keyword.
- Audience and market (India tier-2/3, international, languages).
- Page type: blog / pillar / spoke / service / comparison / FAQ.
- Word count target (default 800–1500 for blogs, 600–1000 for service).
- Tone: simple business (default), premium/corporate, or casual.
- Brand name, product name, and CTAs to include.
- Internal pages on the site that the post should link to (URL list).
- Banned items: words, claims, competitor names.
Ask at most 2 clarifying questions if topic or audience is unclear.
Required resources
- Keyword input from
leadup-keyword-competitor-researcherif available (target keyword, intent, target page). - SERP scan of the top 3–5 ranking pages for the target keyword: their H2s, word count band, schema, FAQs.
- PAA / autocomplete for FAQ questions.
- Brand guidelines if the user provides them.
- Official docs for any technical claim (product, API, regulation).
- Final polish:
leadup-human-content-editorfor AI-style cleanup if the draft starts to feel templated.
Internet research workflow
- Confirm intent of the target keyword (info / commercial / transactional).
- Pull top 3–5 SERP results; record H1, H2 list, word count band, schema.
- Pull PAA box + autocomplete for FAQ candidates (3–6 questions).
- Verify any claim with an official source if technical or regulated (medical, financial, legal); note source.
- Note assumptions and any claim that needs the user's confirmation.
If a browser or search MCP is available, hand off to
leadup-mcp-tool-orchestrator.
Step-by-step workflow
- Restate topic, audience, target keyword, intent, word count.
- Brief (one block): goal, target keyword, secondary keywords, audience, tone, CTAs, must-include facts, must-avoid items.
- Outline (SEO-aware): H1 (matches intent), 4–8 H2s, H3s where needed, FAQ section, CTA placement.
- Title options: 5 titles (≤60 chars), each with a different angle (how-to, list, comparison, mistake, case).
- Meta description: 150–160 chars, target keyword + benefit + CTA.
- Draft the body section by section, following the outline:
- Hook in the first 2 lines.
- Vary sentence length; no em dashes (unless user asks).
- No banned AI hype words.
- Use specific examples (Indian or international, matching market).
- FAQ: 3–6 Q&A pairs derived from PAA; answers 40–80 words each.
- Internal links: 3–7 suggestions from the user's site (anchor + target URL + reason).
- Final pass: run the draft through the human writing rules; if AI
tells remain, route to
leadup-human-content-editor.
Full framework: references/blog-writing-framework.md.
Brief shape: assets/blog-brief.template.md.
Required output format
One Markdown document with these sections, in this order:
- Brief — topic, target keyword, intent, audience, tone, word target.
- Title options (5) — each ≤60 chars, different angle.
- Meta description — single line, 150–160 chars.
- Outline — H1 + H2s + H3s + FAQ + CTA placement.
- Draft — full body in Markdown, headings preserved, no em dashes, no banned hype words, sentence rhythm varied.
- FAQ — 3–6 Q&A pairs.
- Internal links — anchor → URL → reason.
- Assumptions and verification needs — sources used, any claim that needs user confirmation.
Defer final stylistic polish to leadup-human-content-editor if needed.
Safety rules
- Do not use em dashes unless the user explicitly asks.
- Do not use the LeadUp banned hype words (seamless, empower, unlock, robust, cutting-edge, leverage, elevate, harness, supercharge, revolutionize, game-changing, world-class, next-generation, holistic, synergy) unless explicitly required by the brand.
- Do not invent statistics, citations, case studies, client names,
awards, or quotes. If a stat is needed, mark
[verify: source]. - Do not copy competitor content; reference structure only.
- Do not make medical, legal, or financial claims without an official source — flag for user verification.
- Keep prices, dates, and product names exact from the source.
- For India audience: clear international English; avoid Indian-English habits ("do the needful", "kindly revert") unless the brand uses them.
- Respect copyright. Quote sparingly with attribution.
- Defer integrations to
leadup-api-research-builder; defer browser / search tooling toleadup-mcp-tool-orchestrator; defer final polish toleadup-human-content-editor.
Common mistakes
- Hero paragraph that says nothing (filler before getting to the answer).
- Every H2 the same length and shape (AI rhythm).
- Em dashes sneaking back in for "flow".
- Inventing a statistic to make the intro feel authoritative.
- Long bullets where every line has the same verb form.
- Forgetting the FAQ — that's where most informational ranking lives now.
- Stuffing the target keyword in every H2.
- Vague benefits ("boost your business") with no concrete outcome.
- Treating a service page like a blog — service pages need pricing, process, and a strong CTA, not a 1500-word essay.
Troubleshooting
- No keyword research available: pick the most obvious target from the
topic and flag it as estimated; recommend a follow-up research pass
with
leadup-keyword-competitor-researcher. - User wants ultra-short post (200–400 words): drop the FAQ, keep one CTA, one H2 maximum.
- User wants long pillar (3000+ words): split into clear H2s, include a TOC, add 6+ FAQ entries.
- Regulated topic (medical, finance, legal, real estate): flag every claim for user verification; do not state outcomes as guarantees.
- Translation request: do not translate inside this skill; flag and route to a translator/translation pass.
- Brand voice is highly casual / Gen-Z: ask for 2 reference posts and match their pattern instead of defaulting to neutral business tone.
Test prompts
Should trigger (5)
- "Write an SEO blog on 'WhatsApp reminders for clinics'."
- "Draft a service page for our salon booking SaaS."
- "Write a human-style article on 'reduce no-shows in dental clinics'."
- "Write a comparison post: LeadUp Booking vs Calendly for Indian salons."
- "Blog post on 'best invoicing for tier-2 city retailers' — no AI vibes."
Should NOT trigger (3)
- "Find me 100 keywords on dental marketing." (→
leadup-keyword-competitor-researcher) - "Plan 30 days of social posts." (→
leadup-content-calendar-builder) - "Audit our site's SEO." (→
leadup-seo-strategist)
Functional test cases (2)
- Given target keyword "salon software pricing in India" (commercial
intent), return brief + 5 title options ≤60 chars + 150–160 char meta
- outline with FAQ + 800–1200 word draft with zero em dashes, zero banned hype words, and 3+ internal link suggestions.
- Given a regulated topic ("teeth whitening at home" for a dental
clinic site), return a draft that flags every medical claim with
[verify: source], avoids guarantee language, and includes an "Assumptions and verification needs" block listing what the clinic must confirm before publishing.
Success criteria
- Output has all 8 required sections in order.
- Draft has zero em dashes (unless user asked) and zero banned hype words.
- Title options under 60 chars; meta under 160 chars.
- FAQ is derived from real PAA / autocomplete signals, not invented.
- Internal links point to plausible user-site URLs with reasoning.
- Every statistic, citation, or claim is either sourced or marked
[verify: source]. - Tone matches LeadUp's simple, trustworthy, business-specific voice; if
it doesn't, the user is told to run
leadup-human-content-editor.