second-state/market-research-grok/skill

Claude Code and agent skill by second-state/market-research-grok

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Install

npx vskill@latest install second-state/market-research-grok/skill

SKILL.md

Market Research via Grok — X/Web Sentiment Analysis + Product Visuals

Analyze market demand for a product idea by searching X (Twitter) and the web for real user opinions, pain points, and competitive signals. Generates an honest market assessment plus 5 AI-rendered product concept images and 5 animated videos.

Binary

  • {baseDir}/scripts/market_research — Market research CLI.

Prerequisites

  • GROK_API_KEY environment variable set with a valid xAI API key.

Usage

Full run (research + images + videos)
shell
GROK_API_KEY="$GROK_API_KEY" {baseDir}/scripts/market_research \
--product "PRODUCT_DESCRIPTION" \
--output /tmp/market-report.json
Research only (skip media generation)
shell
GROK_API_KEY="$GROK_API_KEY" {baseDir}/scripts/market_research \
--product "PRODUCT_DESCRIPTION" \
--skip-media \
--output /tmp/market-report.json
Parameters
ParameterRequiredDescription
--productYesProduct description: features, user profiles, price points, use cases
--outputNoOutput file path (defaults to stdout)
--termsNoNumber of search terms to generate, 10-20 (default: 15)
--skip-mediaNoSkip image and video generation (research only)
Output

A JSON report containing:

  • product_summary — The input product description
  • search_terms — Generated multi-dimensional search terms
  • findings[] — Per-term analysis with positive/negative signals, quotes, and sentiment
  • synthesis — Final assessment with market need score (1-10), pain points, risks, and opportunities
  • media[] — 5 product concept images and 5 animated videos (unless --skip-media)

Each media entry contains:

  • description — What the image/video shows
  • image_prompt / video_prompt — The prompts used for generation
  • image_url — URL to the generated product concept image
  • video_url — URL to the 5-second animated video
⚠️ Image and video URLs are temporary. Download or present them immediately.

Workflow

1. Gather the Product Description

Ask the user to describe their product idea. Encourage them to include:

  • Key features and differentiators
  • Target user profiles / personas
  • Proposed price points
  • Example use cases or problem scenarios
2. Run the Analysis

For a full analysis with visuals:

shell
GROK_API_KEY="$GROK_API_KEY" {baseDir}/scripts/market_research \
--product "USER_PRODUCT_DESCRIPTION" \
--output /tmp/market-report.json

For quick research without media (faster, cheaper):

shell
GROK_API_KEY="$GROK_API_KEY" {baseDir}/scripts/market_research \
--product "USER_PRODUCT_DESCRIPTION" \
--skip-media \
--output /tmp/market-report.json

The CLI runs six phases:

  1. Term Generation — Grok generates 10-20 search terms covering features, competitors, personas, pricing, and pain points.
  2. Market Research — Each term is searched on X and the web via Grok with live search enabled.
  3. Synthesis — All findings are synthesized into an honest market assessment.
  4. Image Prompts — Grok creates 5 polished image concepts based on the product and synthesis.
  5. Image Generation — Each concept is rendered via grok-imagine-image-pro.
  6. Video Generation — Each image is animated into a 5-second video via grok-imagine-video.
3. Present the Results

Read the JSON output and present a summary to the user:

shell
cat /tmp/market-report.json

Key sections to highlight:

  • Market Need Score (1-10) and description
  • Pain Points Solved vs Pain Points Missed
  • Competitive Landscape
  • Honest Assessment — the frank "would you invest?" verdict
  • Risks and Opportunities
  • Product Visuals — Share the image and video URLs with the user
4. Download Media (Important)

Image and video URLs are temporary. If the user wants to keep them:

shell
# Download all images and videos
cat /tmp/market-report.json | jq -r '.media[].image_url' | xargs -I{} curl -sLO {}
cat /tmp/market-report.json | jq -r '.media[].video_url' | xargs -I{} curl -sLO {}

How It Works

  • Uses Grok (grok-4-0709) for research/synthesis — most advanced model, always reasons at maximum effort
  • Uses grok-imagine-image-pro for images — highest quality ($0.07/image)
  • Uses grok-imagine-video for videos — 5 seconds, 16:9, 720p, image-to-video animation
  • All research calls have X search, web search, and news search enabled
  • The synthesis is designed to be brutally honest, not a cheerleading exercise

Cost Per Run

  • Research only (--skip-media): ~17 Grok API calls (variable with token usage)
  • Full run: research + 5 images ($0.35) + 5 videos ($1.25) = ~$1.60 extra for media

Supported Platforms

  • macOS Apple Silicon (aarch64)
  • macOS Intel (x86_64)
  • Linux x86_64 (static musl)
  • Linux aarch64 (static musl)

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/second-state/market-research-grok/skill/badge)](https://verified-skill.com/skills/second-state/market-research-grok/skill)

Something wrong with this skill or its page? Tell us.