spoonbobo

idea-validator

Validate project ideas before building by scanning GitHub, npm, and PyPI for existing competitors — returns a reality_signal score (0-100) indicating how crowded the space is.

spoonbobo 0 Updated 2mo ago

Resources

1
GitHub

Install

npx skillscat add spoonbobo/trinity/idea-validator

Install via the SkillsCat registry.

SKILL.md

Idea Validator

Scan real data sources before building anything new. Returns a reality_signal score (0-100) indicating how crowded the space is, plus top competitors and pivot suggestions.

When to Activate

IMPORTANT: Run this automatically before starting any new project, feature, or tool.

Also trigger when the user:

  • Says "build me a [tool/app/service]"
  • Asks "is there already a [thing] out there?"
  • Wants to validate a project idea or check for competitors
  • Mentions starting a new side project or product

Usage

uv run {baseDir}/scripts/validate_idea.py --query "AI code review tool"

Options

Flag Description Default
--query / -q The idea or project description to validate (required)
--depth / -d Search depth: quick (GitHub only) or deep (GitHub + npm + PyPI) deep

Example Output

{
  "query": "AI code review tool",
  "reality_signal": 82,
  "verdict": "Very crowded space — differentiate or reconsider",
  "sources": {
    "github": { "total_repos": 1847, "top": [...] },
    "npm": { "total_packages": 23, "top": [...] },
    "pypi": { "total_packages": 15, "top": [...] }
  },
  "top_competitors": [
    { "name": "reviewdog/reviewdog", "stars": 9104, "description": "Automated code review tool", "source": "github" },
    { "name": "sourcery-ai/sourcery", "stars": 1523, "description": "AI-powered code reviewer", "source": "github" }
  ],
  "pivot_hints": [
    "Focus on a specific language or framework",
    "Target a niche industry (finance, healthcare, compliance)",
    "Build for a specific workflow (PR review, pre-commit, IDE integration)"
  ]
}

Decision Thresholds

After running the validator, follow these rules:

Score Action
> 70 (High) STOP. Report top 3 competitors with star counts. Ask the user if they want to proceed, pivot, or abandon.
30–70 (Medium) Show results and pivot hints. Suggest a niche angle that existing projects don't cover.
< 30 (Low) Proceed to build. Mention that the space is open.

Always show the reality_signal score and top competitors before writing any code.

Interpreting Results

  • High signal (>70) doesn't mean "don't build" — it means "differentiate or don't bother." The top players are likely mature and well-funded.
  • Low signal (<30) means genuine white space. Solo builders have the best odds here.
  • Medium signal (30-70) is where nuance matters — look at the pivot hints for angles existing projects miss.

Variations

  • Batch validation: Before a hackathon, validate a list of ideas and rank by reality_signal — lowest score = most original opportunity.
  • Quick mode: Use --depth quick for a fast GitHub-only check when you just need a rough signal.
  • Pre-build gate: Add this to your workflow so it runs automatically before any "build me X" request.

Tips

  • The reality_signal is based on real data (repo counts, star distributions, package existence), not LLM guessing.
  • Public APIs have rate limits — GitHub allows 10 requests/min unauthenticated. For heavy use, set a GITHUB_TOKEN env var.
  • A project with 50k+ stars in the space is a strong signal — but it also validates the market.
  • The script uses only public APIs with no authentication required for basic usage.