name: landing-page-architect
description: Create, audit, or rewrite product and service landing pages using offer diagnosis, bottleneck-aware structure, concrete block-by-block copy, proof, risk reduction, decision details, and CTA guidance. Use when Codex is asked to write landing-page copy, choose a landing-page structure, diagnose conversion issues, improve hero/value proposition/CTA/proof/FAQ/pricing blocks, or create product, service, SaaS, AI, developer-tool, startup, lead-gen, demo-request, waitlist, consultation, local-service, or high-ticket offer pages.
Landing Page Architect
Landing Page Architect is a Codex-compatible AI agent skill for landing page copywriting, conversion-focused page structure, product and service page audits, value proposition writing, CTA improvement, proof/risk-reducer design, and offer-to-page transformation.
Use this skill to create, audit, or rewrite conversion landing pages. Prefer specific offer logic and block copy over generic CRO advice.
Core Rule
Build every landing page around the beliefs the visitor needs before acting:
Orient: I understand what this is.Match: This is for me.Demonstrate: I can picture how it works.Prove: I believe the claim.De-risk: I feel safe acting.Convert: I know the next step.
Read references/playbook.md before writing, auditing, or restructuring a page.
Workflow
- Extract or ask for the minimum offer brief: offer, audience, primary action, outcome, mechanism, proof available, and commitment/risk. If the user wants immediate output and facts are missing, proceed with explicit placeholders instead of inventing details.
- Identify the offer type: product, service, SaaS, AI tool, developer tool, physical product, professional service, local service, high-ticket service, waitlist, trial, demo, lead magnet, consultation, quote request, or marketplace side.
- Identify the primary conversion action: buy, start trial, book call, request quote, join waitlist, request access, subscribe, download, submit inquiry, or apply.
- For a new page, use the complete first-version structure from the playbook, then adapt it for offer type, commitment level, and proof available.
- For an existing page, diagnose the main observable bottleneck:
- Fast exits, low scroll, no CTA engagement: orientation failure.
- Engagement with demos/examples/process before action: tangibility gap.
- CTA/form/pricing/FAQ engagement but abandonment: commitment friction.
- Write each block one by one. For each block, include the block goal, draft copy, required asset/proof if any, and missing input if the user has not supplied it.
- Use one primary CTA. Allow a secondary CTA only when it reduces friction without competing with the main action.
- Separate proof from claims. Use the strongest honest evidence available; do not invent metrics, customers, testimonials, guarantees, pricing, or certifications.
Output Preferences
For creation tasks, provide:
- Recommended structure
- Draft copy for each block
- Missing inputs or assumptions
- Notes on needed proof, visuals, pricing, risk reducers, or CTA clarity
For audit tasks, provide:
- Bottleneck diagnosis
- Highest-impact fixes first
- Block-by-block rewrite guidance
- What to remove or shorten
For first-version copy tasks, use this compact output shape:
Assumptions:
Structure:
Block Drafts:
Missing Assets / Proof:
Highest-Priority Fix:
Constraints
- Keep copy concrete, scannable, and non-hypey.
- Prefer product screenshots, sample outputs, deliverables, process artifacts, case proof, and real examples over abstract claims.
- For services, make process, deliverables, fit, credibility, and next step explicit.
- For products, make product reality, benefits, details, pricing/commitment, support, and risk reducers explicit.
- For high-ticket or sensitive offers, move proof, process, pricing expectations, and risk reducers higher.
- For new categories, AI tools, technical tools, and abstract services, move examples and mechanism higher.