Skip to content
Back to skills

Create An Asset

ASecurity

Generate tailored sales assets (landing pages, decks, one-pagers, workflow demos) from your deal context. Describe your prospect, audience, and goal, get a polished, branded asset ready to share with customers.

  • 65 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 4, 2026
ai-agentsgonode

Security analysis

A100/100

Pro scans all 9 files and shows the line behind each finding

Scanned September 4, 2026

npx -y skills add naveedharri/benai-skills --skill create-an-asset --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Create An Asset?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Create An Asset
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/naveedharri-create-an-asset/badge)](https://www.skillsdirectory.com/skills/naveedharri-create-an-asset)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: create-an-asset
description: Generate tailored sales assets (landing pages, decks, one-pagers, workflow demos) from your deal context. Describe your prospect, audience, and goal, get a polished, branded asset ready to share with customers.
disable-model-invocation: true
---

# Create an Asset

Generate custom sales assets tailored to prospect, audience, and goal. Four formats: interactive landing page, deck-style, one-pager, workflow/architecture demo. Works for any seller, any product, any prospect.

## Triggers

Invoke this skill when:
- User says `/create-an-asset` or `/create-an-asset [CompanyName]`
- User asks to "create an asset", "build a demo", "make a landing page", "mock up a workflow"
- User needs a customer-facing deliverable for a sales conversation

## Steps

### Step 1: Collect context

Read `references/context-collection.md`. Detect seller context from the user's email domain, then collect the four inputs: (a) prospect, (b) audience, (c) purpose, (d) format. Parse what the user already gave before asking. If the format is a workflow demo, also collect components, flow, human touchpoints, and an example scenario (covered in the same file).

### Step 2: Research the prospect

Read `references/research.md`. Assess context richness (rich / moderate / sparse) and run the matching research pass: prospect basics, leadership, and brand colors always; industry, tech stack, and competitive context when context is moderate or sparse; conversation analysis when transcripts or materials were uploaded.

### Step 3: Decide the structure

Read `references/format-selection.md`. Pick sections, slides, or demo layers for the chosen format based on purpose and audience. If the format is a workflow demo, also read `references/workflow-demo.md` for component definitions, flow steps, and scenario narrative.

### Step 4: Ask clarifying questions (required)

Read `references/clarifying-questions.md`. Summarize your plan back to the user, ask the standard questions plus the format-specific ones, and confirm before building. Max 2 rounds; if still ambiguous, make a reasonable choice and say so.

### Step 5: Generate content

Read `references/content-templates.md`. Write every section using the templates and the general principles (their pain points, their language, explicit product-to-need mapping, proof points). For workflow demos, use `references/workflow-demo.md` for components, steps, and narrative.

### Step 6: Apply visual design

Read `references/visual-design.md`. Apply the color system (prospect brand colors by default), typography, and element styles. For workflow demos, `references/workflow-demo.md` has the node, arrow, canvas CSS and icon mappings.

### Step 7: Build, deliver, iterate

Read `references/delivery.md`. Build a single self-contained HTML file, run the quality checklist, deliver with the delivery message (summary, deployment options, customization offers), then handle iteration requests per the iteration table.

## Self-improvement

This skill is never finished. Improve it as you use it.

- When the user corrects how a step was done, update the relevant reference file (or this SKILL.md) so the correction sticks. Do not just fix it for this run.
- When a correction is a hard rule ("always do X", "never do Y"), add it as a permanent rule here.
- When the user says an output was genuinely good, save it to `references/examples/` so it becomes a model for future runs.
- Keep the skill small while you do this: when you add something, run the deletion test and cut anything that no longer changes behavior.

Files in this skill

  • SKILL.md3.5 KB
  • references/clarifying-questions.md2.2 KB
  • references/content-templates.md2.1 KB
  • references/context-collection.md5.2 KB
  • references/delivery.md3.4 KB
  • references/format-selection.md5.3 KB
  • references/research.md2.1 KB
  • references/visual-design.md1.8 KB
  • references/workflow-demo.md3.7 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.