agentleFS
jcottamjcottam/agent-resources/skills/design/content-wireframe/SKILL.md

content-wireframe

Skill25 starsChanged 6 months agoMITView on GitHub

Draft a content wireframe for a website or page before any design or code. Produces a structured markdown document with real copy, section intent, and open decisions — approved by the user before implementation begins. Use when the user says "content first", "wireframe the copy", "words before design", "start with the content", "what should the page say", or when a website project lacks finalized copy and is jumping straight to visuals.

The file

Show its rules and text

We split this skill into the rules it gives an agent

20 rules. About 0 are common, 20 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈20

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write every word a page needs — in order, with intent annotations — before touching layout, color, or code.

    line 14
  2. Distinctive:

    Before drafting, collect as much of the following as possible:

    line 25
  3. Distinctive:

    After gathering inputs, define the audience before writing a word.

    line 40
  4. Distinctive:

    Before drafting, establish the page's voice concretely — not just "casual" or "professional," which mean different things to different …

    line 55
  5. Distinctive:

    Write 3 sample sentences in the proposed voice, covering different tones the page will need (e.g., a hero headline, an explanatory …

    line 58
  6. Distinctive:

    Write all copy in the calibrated voice.

    line 63
  7. Distinctive:

    If competitor pages are available (from inputs or by crawling), analyze them before drafting:

    line 68
  8. Distinctive:

    Use this to sharpen the wireframe's positioning — lean into the gaps

    line 74
  9. Distinctive:

    Every section must have an intent annotation and real copy.

    line 87
  10. Distinctive:

    Not all sections fit the headline → copy → CTA mold.

    line 120
  11. Distinctive:

    Real words only: Never use placeholder text.

    line 195
  12. Distinctive:

    No design decisions: No colors, fonts, sizes, layouts, components, or code.

    line 197
  13. Distinctive:

    Every section earns its place: If a section doesn't serve the primary goal or build toward conversion, question whether it belongs on this …

    line 199
  14. Distinctive:

    Intent before content: Write the intent annotation first.

    line 201
  15. Distinctive:

    Flag, don't assume: When you hit a decision that depends on the user's judgment (brand tone, pricing strategy, what photo to use), surface …

    line 203
  16. Distinctive:

    Consolidate proof points by default: Prefer grouping trust signals so the case is made once, clearly.

    line 206
  17. Distinctive:

    Read all available context (project docs, existing pages, data files, brand guides)

    line 218
  18. Distinctive:

    Write the competitive landscape summary (if applicable)

    line 225
  19. Distinctive:

    Write real copy for every heading, paragraph, and CTA

    line 227
  20. Distinctive:

    Do not proceed to implementation until the user explicitly approves the copy.

    line 238

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Content Wireframe

Write every word a page needs — in order, with intent annotations — before touching layout, color, or code. The wireframe is the blueprint; design follows.

When to use

  • New website or landing page where copy hasn't been written
  • Redesign where the existing page was built design-outward and the words feel secondary
  • Any project where the user is unhappy with the "feel" but can't articulate why (often a content problem, not a design problem)

Inputs to gather

Before drafting, collect as much of the following as possible:

  1. Business context — What does the company do? Who are the customers?
  2. Primary goal — What is the single most important action a visitor should take?
  3. Brand voice — Formal/casual? Technical/approachable? Any existing style guide?
  4. Existing copy — Current site, marketing materials, internal docs
  5. Competitor pages — Direct competitors' sites, their messaging, positioning, and gaps
  6. Data files — Locations, pricing, FAQs, testimonials, team bios — anything structured
  7. Constraints — Pre-launch? Pricing TBD? Missing photography? Legal requirements?

Sources: AGENTS.md, README.md, existing page files, data modules, brand guides, and conversation context.

Audience model

After gathering inputs, define the audience before writing a word. Produce a short profile for each distinct audience segment the page must address:

  • Who they are — Role, situation, level of awareness of the problem/product
  • What they already believe — Assumptions, prior experiences, mental models
  • What objections they carry — Reasons they might leave without acting
  • What language they use — How they describe the problem in their own words

Reference the relevant audience segment in each section's intent annotation (e.g., "This section targets first-time visitors who don't yet understand what the product does"). If a section doesn't clearly serve at least one segment, question whether it belongs.

Voice calibration

Before drafting, establish the page's voice concretely — not just "casual" or "professional," which mean different things to different people.

  1. Write 3 sample sentences in the proposed voice, covering different tones the page will need (e.g., a hero headline, an explanatory paragraph, an FAQ answer).
  2. Present these to the user for confirmation. Adjust until the voice feels right.
  3. Write all copy in the calibrated voice. If voice drifts between sections, catch it in review.

Competitive messaging analysis

If competitor pages are available (from inputs or by crawling), analyze them before drafting:

  1. Note each competitor's headline, value proposition, and primary CTA
  2. Identify messaging patterns — what claims do they all make?
  3. Find gaps — what do competitors fail to say, underemphasize, or get wrong?
  4. Use this to sharpen the wireframe's positioning — lean into the gaps

Summarize findings in a short "Competitive landscape" section at the top of the wireframe document, before the page sections begin. This gives the user context for why certain messaging choices were made.

Output format

Produce a single markdown file (e.g., HOMEPAGE-WIREFRAME.md) that reads top to bottom like a visitor scrolling the page.

Section template

Every section must have an intent annotation and real copy. The structure below each intent adapts to the content type — not every section is headline + paragraph.

## [N]. [Section Name]

> **Intent:** What is this section's job? What should the visitor understand,
> feel, or do after reading it? Which audience segment does this serve?
>
> **Density:** [Light | Medium | Dense] — see guidance below

### Headline

[The actual headline words]

### Copy

[The actual paragraphs, bullets, labels — real words, not lorem ipsum]

### CTA

[Exact button/link text] → [destination]

> **Open question:** [Any decision the user needs to make — phrasing alternatives,
> missing info, things to confirm with stakeholders]
>
> **Content note:** [Constraints, dependencies, or context that affects this section]

Omit sub-sections that don't apply (e.g., not every section has a CTA).

Variant structures

Not all sections fit the headline → copy → CTA mold. Use the appropriate structure for the content:

  • Data sections (pricing, specs, comparisons): Use tables or structured lists. Write column headers, row labels, and footnotes as real copy.
  • Multi-item sections (features, team members, locations): Write the section-level headline and intro, then each item as a named entry with its own copy.
  • Media-anchored sections (hero image, video, demo): Describe the media asset needed (subject, mood, what it should communicate) and write any overlay text, captions, or surrounding copy.
  • Interactive sections (calculators, configurators, maps): Describe the interaction, write all labels and microcopy, and note what inputs/outputs the user sees.

Copy density guidance

Match density to the section's job:

DensityWhen to useTypical length
LightHero, closing CTA, navigation — sections that rely on a few precise words1-2 sentences, a single headline, or just labels
MediumValue proposition, how-it-works, about — sections that explain or persuade1-3 short paragraphs or a headline + supporting bullets
DenseFAQ, pricing details, product specs — sections where visitors come to get answersAs long as needed, but structured with headers, tables, or lists

Sections to consider

Starting checklist — not a fixed template. Adapt aggressively based on the page type. A SaaS landing page and a local restaurant homepage share almost none of these in equal proportion.

Load-bearing sections — these carry the conversion argument. Spend the most drafting effort here:

SectionJob
Hero / OpeningIdentify + hook — answer "what is this?" in 5 seconds
Value propositionWhy this company over alternatives
Closing CTAFinal conversion or connection push

Supporting sections — these build toward conversion but don't carry it alone. Include based on what the audience needs:

SectionJob
NavigationOrient visitors, surface the primary action
How it worksReduce anxiety about the process
Product / Service detailsInform the purchase decision
Social proofBuild trust (testimonials, ratings, logos, history)
FAQCatch remaining objections
Story / AboutEmotional connection and differentiation

Structural sections — necessary but rarely decisive:

SectionJob
Locations / Contact infoLogistics and reachability
FooterReference info, legal, secondary links

End with collected decisions

Close the wireframe with a numbered list of every open question and content decision surfaced throughout the document. This gives the user a single checklist to work through.

## Open decisions (collected)

1. **[Topic]:** [Question or choice to resolve]
2. **[Topic]:** [Question or choice to resolve]
...

Principles

  • Real words only. Never use placeholder text. If you don't know the right words yet, write your best guess and flag it as an open question.
  • No design decisions. No colors, fonts, sizes, layouts, components, or code. The wireframe is a reading experience, not a visual one.
  • Every section earns its place. If a section doesn't serve the primary goal or build toward conversion, question whether it belongs on this page.
  • Intent before content. Write the intent annotation first. If you can't articulate what a section's job is, it probably shouldn't exist.
  • Flag, don't assume. When you hit a decision that depends on the user's judgment (brand tone, pricing strategy, what photo to use), surface it as an open question rather than making the call silently.
  • Consolidate proof points by default. Prefer grouping trust signals so the case is made once, clearly. Exception: when a specific feature or claim benefits from an adjacent testimonial or data point, place it there and note why in the intent annotation.
  • SEO is out of scope. This skill produces copy optimized for the reader, not for search engines. If keyword placement matters, flag it as an open decision for the user to address separately or with an SEO-focused tool.

Workflow

Phase 1: Research

  1. Read all available context (project docs, existing pages, data files, brand guides)
  2. Build the audience model (see above)
  3. Analyze competitor messaging if competitor pages are available
  4. Calibrate voice with sample sentences and get user confirmation

Phase 2: Draft

  1. Write the competitive landscape summary (if applicable)
  2. Draft the wireframe following the section template above
  3. Write real copy for every heading, paragraph, and CTA
  4. Annotate each section with intent, audience segment, and density
  5. Flag open decisions inline
  6. Collect all open decisions into the closing checklist

Phase 3: Review

Present the wireframe to the user. They read it, react, rewrite sections, and resolve open decisions. Iterate until every word feels right and every section earns its place.

Do not proceed to implementation until the user explicitly approves the copy.

Handoff

Once approved, the wireframe is the spec for implementation. Hand it to the design or implementation skill of your choice. The wireframe itself makes no layout, visual, or code decisions — that's the implementer's job.

Anti-patterns

  • Writing copy to fill a design. The whole point of this exercise is to avoid that. If you catch yourself thinking about column counts or card layouts while writing copy, stop.
  • Skipping intent annotations. These are the most valuable part. They force clarity about why each section exists.
  • Burying open questions. Surface them inline AND in the collected list at the end. The user should never be surprised by an assumption you made.
  • Perfectionism in the first draft. The draft is meant to be reacted to, not shipped. Write your best guess and let the user correct it.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.