Landing Copy

Pack Availability Guard

Before telling the user to run a skill from another project-local pack, check .agents/project.json.enabled_packs. If the target pack is not enabled, recommend $pack install <pack> instead of the target skill. Global skills are always valid. Skills from this same pack are valid because the current skill is already running from that pack.

Landing Copy — Research-Grounded Landing Page Copy

Report-First Approval Gate

Default to report-only: present findings, evidence coverage, assumptions, recommended artifact path, and proposed file changes in a pre-approval alignment page plus a concise conversation summary for user approval before creating or updating canonical research, spec, or task files.

Do not write or overwrite synthesized deliverables until the user explicitly approves, unless the user invoked an explicit write/update/fix mode or clearly asked to write files upfront. Raw evidence capture may be persisted before analysis when reproducibility requires it; report those raw paths separately and still gate synthesized research/report writes.

When stopping for approval, build and attempt to open the alignment preview page first, then ask the user to review it and approve, question, or request adjustments. Do not include Recommended next skill, Recommended next command, or downstream routing language. The approval request itself is the next action. Only emit next-skill routing after the approved artifact has been written or updated.

Dual-mode skill that either generates complete landing page copy from upstream research artifacts or audits existing copy against codebase reality, research drift, and conversion best practices. Every claim in the generated copy links back to a research source so nothing is hand-waved.

Prerequisites

• Hard: None. Works with just a README or codebase — greenfield projects can use it immediately.

• Soft (enhancement tiers — display the applicable tier to the user before starting):

  Tier	Research Docs Present	Copy Quality
  1	None — only README / codebase	Functional but generic; many claims will be ungrounded
  2	positioning.md + icp.md	Targeted messaging with real pain points and value props
  3	Tier 2 + gtm.md + competitive-analysis.md + journey-map.md	Full funnel copy with differentiation and onboarding flow
  4	Tier 3 + customer-feedback.md + monetization.md + metrics.md	Production-grade copy with social proof, pricing, and data-backed claims

  Read each if it exists (respect {app}/ prefix when monorepo scope is active):

  • research/positioning.md — market category, positioning statement, unique attributes, value mapping
  • research/gtm.md — messaging framework, one-liner, channel strategy, launch plan
  • research/icp.md — ICP segments, pain points, value props, trigger events, customer language
  • research/journey-map.md — onboarding flow, aha moment, task flows
  • research/customer-feedback.md — real customer quotes, recurring themes, objections
  • research/monetization.md — pricing tiers, packaging, free vs paid boundaries
  • research/metrics.md — north star metric, activation/engagement/retention metrics
  • research/competitive-analysis.md — competitor positioning, strengths, weaknesses, market gaps

Process

0. App Scope Resolution (Monorepo Support)

Before anything else, determine the app scope:

1. If $ARGUMENTS specifies an app name matching a subdirectory of research/, use it.
2. If research/ contains subdirectories (excluding files), list them and ask the user which app to target. If only one subdirectory exists, use it automatically.
3. If no subdirectories exist, proceed with flat structure (single-product mode).

When app scope {app} is active:

• Read/write research from research/{app}/ instead of research/

1. Determine Mode

Parse $ARGUMENTS:

• generate (default): Create complete landing page copy from research + codebase context
• audit: Check existing copy against codebase reality, research drift, and conversion best practices

If mode is audit, jump to the Audit Mode section below.

Generate Mode

2. Load Context

Read all available research docs listed in Prerequisites (soft). For each:

• Note which docs exist and which are missing
• Extract the key claims, data points, and customer language that will ground copy sections

Also read:

• CLAUDE.md, README.md — product name, description, core value proposition
• Key source files (routes, components, package.json) — to verify feature claims against reality

Determine the enhancement tier and display it to the user:

Research coverage: Tier [N] — [tier description from table above]. Missing docs that would improve copy: [list missing soft prerequisites with their $skill commands].

3. Web Research

Use WebSearch with 4–6 category-aware queries:

1. Landing page best practices — "SaaS landing page copy best practices [year]", "high-converting landing page structure"
2. Category-specific examples — "[market category from positioning.md] landing page examples", "best [category] homepage copy"
3. Hero section patterns — "above the fold copy patterns", "hero section conversion optimization"
4. Social proof patterns — "social proof landing page examples", "trust signals [category]"
5. CTA optimization — "call to action best practices SaaS", "pricing page CTA patterns"
6. Objection handling — "FAQ section landing page", "reducing purchase anxiety [category]"

4. Interview

If the session is already in Plan mode, prefer request_user_input; otherwise ask in plain text.

Brand Voice & Tone

• "What tone should the copy strike? (e.g., professional but approachable, bold and provocative, calm and reassuring)"
• "Any words or phrases you always use? Any you'd never use?"
• "Show me 1–2 examples of copy you admire (competitors, other products, anything)."

Claims Boundaries

• "What claims can we make confidently? (e.g., 'saves 10 hours/week' — do you have data for this?)"
• "Any claims we must NOT make? (legal, regulatory, unverified metrics)"
• "Do you have specific numbers we can cite? (users, uptime, speed benchmarks, customer count)"

Social Proof Availability

• "Do you have customer quotes or testimonials we can use? Company logos?"
• "Any press mentions, awards, or notable integrations?"
• "Case study data? (before/after metrics, customer stories)"

CTA Strategy

• "What's the primary CTA? (e.g., 'Start free trial', 'Book a demo', 'Get started')"
• "Secondary CTA? (e.g., 'See pricing', 'Watch demo', 'Read docs')"
• "Any urgency or scarcity elements? (limited beta, launch pricing)"

Structural Constraints

• "Any sections you definitely want or definitely don't want?"
• "Target audience for this page — is it the primary ICP or a broader audience?"
• "Any existing brand guidelines, style guides, or design constraints?"

5. Generate Sections (Parallel Subagents)

Launch 10 parallel subagents, one per section. Each subagent receives the full research context and interview answers, and produces copy for its section with inline source attribution.

Section assignments:

1. Hero — Headline, subheadline, primary CTA, supporting visual description. Sources: positioning.md (positioning statement, market category), gtm.md (one-liner, messaging framework).

2. Problem Agitation — Articulate the pain the audience feels today. Sources: icp.md (pain points, trigger events), customer-feedback.md (customer language, recurring complaints).

3. Solution Introduction — Introduce the product as the answer to the pain. Sources: positioning.md (market category, unique attributes), icp.md (value props).

4. Benefits — 3–5 benefit blocks (not features — outcomes). Sources: positioning.md (value mapping), icp.md (value props per segment).

5. How It Works — 3–4 step walkthrough from signup to value. Sources: journey-map.md (onboarding flow, aha moment, task flows).

6. Social Proof — Testimonials, logos, metrics, trust signals. Sources: customer-feedback.md (quotes, themes), metrics.md (north star, key numbers), interview answers.

7. Use Cases / Audience — Who this is for, with scenario-specific hooks. Sources: icp.md (user profiles, segments), journey-map.md (use-case task flows).

8. Pricing — Tier summary, value anchoring, CTA per tier. Sources: monetization.md (pricing tiers, packaging, free vs paid).

9. FAQ / Objection Handling — Address top objections and common questions. Sources: competitive-analysis.md (competitor strengths to counter), customer-feedback.md (objections, concerns).

10. Final CTA + Footer — Closing argument, urgency, secondary links. Sources: gtm.md (launch strategy, urgency elements), interview CTA answers.

Each subagent must:

• Write copy for its section only
• Add <!-- Source: research/[file].md — [specific section/claim] --> comments after each grounded claim
• Flag any claim that has NO research backing as <!-- UNGROUNDED: [claim] — no research source available -->
• Respect brand voice and claims boundaries from the interview

6. Assemble and Present

Combine all 10 sections into a complete landing page copy document. Present to the user with:

• The full copy, section by section

• A Research Grounding Summary table:

  Section	Grounded Claims	Ungrounded Claims	Primary Sources
  Hero	N	N	positioning.md, gtm.md
  ...	...	...	...

• If any section has >50% ungrounded claims, flag it: "⚠ [Section] has limited research backing — consider running [$skill] first."

If the session is already in Plan mode, prefer request_user_input; otherwise ask in plain text: "Ready to write this to research/landing-copy.md? Any sections to revise?"

7. Write Output

Only after user confirms, write:

• research/landing-copy.md (or research/{app}/landing-copy.md) — the full copy with source attribution
• research/landing-copy-interview.md (or research/{app}/landing-copy-interview.md) — interview Q&A log

8. Downstream Impact Check

Landing copy is a terminal artifact — it doesn't feed into other research skills. Perform a lightweight check:

• If research/gtm.md exists and has a ## Messaging Framework section, compare the hero headline and one-liner against it. Flag if they've diverged.
• If research/positioning.md exists, confirm the positioning statement is reflected in the hero and solution sections.

Classify impact:

• None: No conflicts. Skip display.
• Minor: 1–2 messaging inconsistencies. Display inline.
• Major: Hero/positioning fundamentally misaligned. Recommend $reconcile-research.

9. Next Steps

Include a Recommended item (the single highest-impact next step given current project state) with a one-line reason, followed by Other options (2–4 alternatives). Use this format in the output:

Next Steps

Recommended: [recommended skill] — [one-line reason why this is the highest-impact next action given current state]

Other options:

• $skill — [description]
• ...

Recommendation priority (first applicable becomes the recommendation):

1. IF no research/positioning.md: recommend $positioning — positioning drives hero copy; the headline is guesswork without it
2. IF no research/customer-feedback.md: recommend $customer-feedback — real customer quotes would replace placeholder social proof
3. Otherwise: recommend "Implement the copy in your codebase — the source attribution comments show where each claim comes from"

Other options (include all applicable items not chosen as recommended):

• IF no research/positioning.md: $positioning — Positioning drives hero copy; running it would significantly strengthen the headline
• IF no research/customer-feedback.md: $customer-feedback — Real customer quotes would replace placeholder social proof
• IF no research/monetization.md: $monetization — Pricing section is thin without a monetization strategy
• IF no research/journey-map.md: $journey-map — "How It Works" section would benefit from mapped onboarding flows
• ALWAYS: "Implement the copy in your codebase — the source attribution comments show where each claim comes from"
• IF research/gtm.md exists: $gtm — Refresh messaging framework to align with the final landing copy

Audit Mode

A1. Locate Existing Copy

Search for landing page copy in order of priority:

1. research/landing-copy.md (or research/{app}/landing-copy.md)
2. src/pages/index.*, src/app/page.*, app/page.*, pages/index.*
3. index.html, public/index.html
4. Any file matching *landing*, *homepage* in common source directories

If no copy is found, tell the user and suggest running $landing-copy generate instead.

A2. Extract Verifiable Claims

Parse the located copy and extract every verifiable claim:

• Feature claims ("supports X", "integrates with Y")
• Performance claims ("10x faster", "99.9% uptime")
• Social proof claims ("trusted by N companies", "rated #1")
• Positioning claims ("the only X that Y", "unlike Z")
• Pricing claims (tier names, prices, feature lists)

A3. Run Audit Categories (4 Parallel Subagents)

Category 1: Staleness vs. Codebase

• Check feature claims against actual codebase (do the features exist? have any been removed?)
• Check integration claims (are the integrations still present?)
• Flag features in codebase NOT mentioned in copy (missed selling points)

Category 2: Drift from Research Docs

• Compare claims against research/positioning.md (has the positioning changed?)
• Compare against research/gtm.md (has the messaging framework changed?)
• Compare against research/icp.md (are we still targeting the same audience?)
• Flag copy that contradicts current research

Category 3: Conversion Best Practices

• Check for missing sections (no hero? no social proof? no CTA?)
• Evaluate CTA strength and placement
• Check for above-the-fold value proposition clarity
• Assess objection handling coverage

Category 4: Broken Promises

• Claims the product can't currently deliver
• Metrics that are outdated or unverifiable
• Social proof that may be stale (old customer counts, outdated quotes)

A4. Classify and Present Findings

Classify each finding:

• Error: Copy contradicts codebase or makes undeliverable claims
• Warning: Copy drifts from research or misses best practices
• Info: Suggestions for improvement, missed opportunities

Present interactively, errors first:

## Audit Results

### Errors (N)
1. **[Section]**: "[exact claim]" — [why it's wrong] (Source: [file:evidence])
...

### Warnings (N)
...

### Info (N)
...

If the session is already in Plan mode, prefer request_user_input; otherwise ask in plain text: "Want me to generate updated copy for any of these sections?"

A5. Read-Only

Audit mode does not write any files. It is purely diagnostic. If the user wants fixes, suggest running $landing-copy generate with the audit findings as context.

Output

research/landing-copy.md (or research/{app}/landing-copy.md)

# Landing Page Copy

> Based on: [list all research docs used]
> Date: [current date]
> Enhancement Tier: [1-4]

## Hero
<!-- Source: research/positioning.md — Positioning Statement -->
<!-- Source: research/gtm.md — One-Liner -->

**Headline**: [headline]
**Subheadline**: [subheadline]
**Primary CTA**: [CTA text] → [CTA destination]
**Supporting visual**: [description of what the hero image/video should convey]

## Problem Agitation
<!-- Source: research/icp.md — Pain Points -->
<!-- Source: research/customer-feedback.md — Customer Language -->

[2-3 paragraphs articulating the pain]

## Solution Introduction
<!-- Source: research/positioning.md — Market Category, Unique Attributes -->

[1-2 paragraphs introducing the product as the answer]

## Benefits

### [Benefit 1 — Outcome-Focused Title]
<!-- Source: research/positioning.md — Value Mapping -->
[Benefit description]

### [Benefit 2]
<!-- Source: research/icp.md — Value Props -->
[Benefit description]

### [Benefit 3]
[Benefit description]

## How It Works

### Step 1: [Action]
<!-- Source: research/journey-map.md — Onboarding Flow -->
[Description]

### Step 2: [Action]
[Description]

### Step 3: [Action — The Aha Moment]
<!-- Source: research/journey-map.md — Aha Moment -->
[Description]

## Social Proof
<!-- Source: research/customer-feedback.md — Quotes -->
<!-- Source: research/metrics.md — Key Numbers -->

[Testimonials, logos, metrics]

## Use Cases

### For [Audience Segment 1]
<!-- Source: research/icp.md — User Profiles -->
[Scenario-specific hook and value]

### For [Audience Segment 2]
[Scenario-specific hook and value]

## Pricing
<!-- Source: research/monetization.md — Pricing Tiers -->

| [Tier 1] | [Tier 2] | [Tier 3] |
|-----------|-----------|-----------|
| [Price] | [Price] | [Price] |
| [Features] | [Features] | [Features] |
| [CTA] | [CTA] | [CTA] |

## FAQ
<!-- Source: research/competitive-analysis.md — Competitor Strengths -->
<!-- Source: research/customer-feedback.md — Objections -->

**Q: [Common objection as question]**
A: [Answer that addresses the objection]

**Q: [Question]**
A: [Answer]

## Final CTA
<!-- Source: research/gtm.md — Launch Strategy -->

[Closing argument]
**[Primary CTA]** | [Secondary CTA]

---

## Research Grounding Summary

| Section | Grounded | Ungrounded | Primary Sources |
|---------|----------|------------|-----------------|
| Hero | N | N | [sources] |
| Problem Agitation | N | N | [sources] |
| Solution | N | N | [sources] |
| Benefits | N | N | [sources] |
| How It Works | N | N | [sources] |
| Social Proof | N | N | [sources] |
| Use Cases | N | N | [sources] |
| Pricing | N | N | [sources] |
| FAQ | N | N | [sources] |
| Final CTA | N | N | [sources] |

### Ungrounded Claims
[List each ungrounded claim and what research would be needed to ground it]

### Missing Research
[List research docs that would most improve copy quality, with $skill commands]

## Next Steps

**Recommended:** `$skill` — [one-line reason]

Other options:
- [conditional items from step 9 — only include items whose conditions are met]

research/landing-copy-interview.md (or research/{app}/landing-copy-interview.md)

Interview Q&A log — brand voice decisions, claims boundaries, social proof inventory, CTA strategy, structural constraints.

Create the research/ directory if it doesn't exist.

Task Classification

When this skill produces follow-up work, file it by execution semantics:

• Immediately actionable implementation or documentation work goes in tasks/todo.md.
• Human-only external actions tied to automated steps go in tasks/manual-todo.md with _(blocks: Step N.X)_ or _(after: Step N.X)_; repo edits, SDK wiring, generated assets, local commands, tests, audits, and authenticated CLI/API work stay in tasks/todo.md.
• One-time condition-gated records, baselines, or future measurements go in tasks/record-todo.md with source, condition, non-blocking reason, evidence, and promotion rule.
• Cadence-based reviews, playtests, adoption checks, investor updates, retros, or docs-health checks go in tasks/recurring-todo.md with cadence, owner/agent, next due, evidence path, and escalation conditions.
• Do not put non-blocking records or recurring obligations in tasks/todo.md unless they have been explicitly promoted into current execution work.

Constraints

• Present before writing. Never write output files until the full copy has been presented and the user confirms.
• Source-attribute every claim. Every factual claim in the copy must have an inline HTML comment citing its research source. If no source exists, mark it UNGROUNDED.
• Respect claims boundaries. Never include claims the user explicitly ruled out in the interview. Never fabricate metrics or social proof.
• Do not overwrite existing research/landing-copy.md without asking the user first.
• Copy ≠ design. This skill produces the words, not the layout. Visual suggestions are limited to "supporting visual" descriptions, not mockups.
• Audit mode is read-only. Never write files in audit mode.
• Brand voice is law. Once the user defines tone and vocabulary in the interview, every section must respect it — no section should feel like it was written by a different voice.
• Check tasks/manual-todo.md — if it exists, read it. If it lists items related to landing page copy, messaging, or marketing, incorporate those requirements into your work and call them out to the user.

Alignment Page

When this skill produces durable deliverables (research, specs, plans, reports, prototypes, or any document output), build a full-depth HTML alignment page at alignment/landing-copy-{topic}.html. Use a normalized topic slug derived from the app, feature, research subject, report subject, or output filename.

Full content requirement. The alignment page must contain the complete content of every proposed markdown deliverable -- every section, every finding, every detail, every list item. It is a thorough interactive review document, not a summary. Render the full deliverable content in clean, readable HTML with appropriate hierarchy, styling, and navigation. If the skill writes multiple scoped deliverables in one run, build one alignment page that contains all deliverables with anchor-linked navigation. Durable tracker artifacts, such as research/assumption-tracker.md, remain canonical markdown outputs but must also be fully rendered into the alignment page before approval.

Dark-mode styling. Use a dark color scheme by default. Base CSS variables: --bg: #0d1117; --surface: #161b22; --border: #30363d; --text: #c9d1d9; --text-muted: #8b949e; --accent: #58a6ff; --green: #3fb950; --red: #f85149; --orange: #d29922; --purple: #bc8cff;. Apply background: var(--bg); color: var(--text); on body. Use --surface for cards, nav, and table headers. Use --border for all borders. Use --purple for question blocks and gate headings. Use --accent for links and section headings. Keep headings color: #fff or var(--accent) for hierarchy. Question block backgrounds should use #1c2333.

Alignment gates. Treat gates as explicit review sections inside the HTML page. A gate blocks finalization until its required inline questions are answered and compiled into YAML. Include every gate that applies to the skill output, and include these gate types whenever relevant: evidence coverage, assumptions/confidence, scope/non-goals, candidate/verdict decisions, artifact destination, proposed file changes, coverage checkpoint, and post-approval route.

Report-only research gates. For report-only or pre-approval research skills, the alignment page must explicitly contain evidence coverage, assumptions/confidence, recommended path, proposed file changes, and approval gates before any canonical research, spec, or task file is created or updated.

Required inline questions. Each gate must contain at least one required inline question placed directly under the content it governs, inside a visually distinct question block. Each question must use radio-button inputs and include two standing options after the skill-generated choices: "Other / None of the above" backed by a multi-line text box for free-form input, and "Need clarification" backed by an optional notes box where the user can explain what is unclear. When any radio option other than "Other" or "Need clarification" is selected, show an optional "Additional notes" text box beneath it so the user can qualify their choice. Generate questions based on what genuinely needs user input -- do not add filler questions. Do not create a separate bottom "Decisions & Clarifications" section.

Gate YAML contract. At the bottom of the page, include a "Compile Answers" button that aggregates answers from all inline gate questions throughout the page, including free-text notes. The button remains disabled until every required question has a selection, shows a count of remaining unanswered questions, and scrolls to the first unanswered question if clicked early. When every question is answered, generate a structured YAML block with one item per gate answer using this stable shape: section, gate_type, status (answered, other, or needs-clarification), answer, optional notes, and optional target_artifact or target_path when the gate controls file output. After successful compilation, automatically attempt to copy the YAML to the clipboard with the Clipboard API, display copy status, and display the YAML in a read-only textarea with an explicit "Copy YAML" button. The copy button must retry clipboard copy when supported and fall back to selecting the textarea contents when clipboard access is unavailable or blocked.

Pre-approval stop. Before user approval, the next action is review of the HTML alignment page, not downstream routing. Ask the user to review the page and provide the compiled YAML answers. Do not include Recommended next skill, Recommended next command, or downstream routing language until after compiled YAML has been provided and the approved artifacts have been written or updated.

Diff highlighting on updates. When the agent updates an existing alignment page after receiving compiled answers, highlight what changed since the previous version. The agent chooses inline annotation or side-by-side layout per situation.

Archiving. Before replacing an existing alignment page, archive it to docs/history/archive/YYYY-MM-DD/HHMMSS/alignment/landing-copy-{topic}.html.

Browser open. Attempt to open the resulting HTML page in the browser and report whether the open succeeded or was blocked. A blocked browser-open attempt does not make the skill fail when the files were written correctly.

Archive-First Replacement Policy

• Before replacing or substantively rewriting an existing canonical research/spec document (research/**/*.md, specs/**/*.md, or docs/specifications/**/*.md), copy the current file to docs/history/archive/YYYY-MM-DD/HHMMSS/<original-relative-path>.
• Preserve the archived snapshot exactly as it existed before the change; do not edit the archived copy after creating it.
• After the archive snapshot exists, write the updated document to the original canonical path.
• Report both the archive path and the updated canonical path in the final output.
• New files do not need archive snapshots. Append-only updates do not need archive snapshots unless an existing section is regenerated or rewritten.
• Keep any existing user approval requirement before overwriting or replacing a document; archiving does not replace asking when the skill already requires approval.

Default Shipping Contract

• Default next-step routing: when reporting completion, include either Recommended next skill: <command> or the two-line pair **Next work:** <specific task or "none"> and **Recommended next command:** <one command or route> so the next operator has a concrete handoff.
• If this skill creates or modifies tracked repository files, finish by committing and pushing all intended changes to the repository primary branch (main when present, otherwise master) before stopping, even if the user did not explicitly ask for commit/push.
• Do not leave tracked changes or unpushed commits behind. If unrelated tracked work is already present, either include it in sensible commits too or stop and explain the blocker.
• This contract does not override stricter safety rules about secrets, destructive history changes, release publication/tag confirmation, or production deploy confirmation.