phtui-idea-pocket

name	phtui-idea-pocket
description	Use when an agent needs low-token Product Hunt inspiration for service ideas, market scanning, feature references, or MVP positioning. Prefer phtui CLI JSON commands over MCP when possible.
version	1.0.0
author	qyinm/phtui
license	MIT
metadata	{"tags":["product-hunt","ideas","cli","inspiration","market-research","agents"]}

phtui Idea Pocket

Overview

phtui is a Product Hunt-powered idea pocket for agents building new services.

Use the CLI first. MCP is available, but CLI JSON output is usually cheaper in tokens because the agent can request a narrow command, inspect only the fields it needs, and avoid loading a broad MCP tool context.

Install

If phtui is already installed:

phtui help

Install latest from Go:

go install github.com/qyinm/phtui@latest

Or run from a checked-out repo:

go run . help

Core Workflow

1. Start broad with ideas.
2. Pick 1-3 products that match the user's product direction.
3. Fetch details only for those selected products.
4. Summarize concrete reusable patterns: target user, core workflow, pricing clue, feature hook, and possible MVP angle.
5. Do not paste full raw JSON unless the user asks. Extract only the relevant signals.

Commands

Category inspiration

Use this when the user has a domain, category, or product area.

phtui ideas --category ai-agents --limit 5

From source checkout:

go run . ideas --category ai-agents --limit 5

Expected JSON shape:

{
  "source": "category",
  "category_slug": "ai-agents",
  "total": 5,
  "items": [
    {
      "product": {
        "slug": "example",
        "name": "Example",
        "tagline": "Short positioning",
        "votes": 123,
        "comments": 8,
        "rank": 1,
        "categories": ["AI Agents"]
      },
      "inspiration_reason": "...",
      "feature_signals": ["..."],
      "market_signals": ["..."]
    }
  ]
}

Leaderboard inspiration

Use this when the user asks what is currently popular or wants broad market scanning.

phtui ideas --period daily --limit 5
phtui ideas --period weekly --limit 10
phtui ideas --period monthly --date 2026-05-01 --limit 10

Product detail

Use this after selecting a product slug from ideas or leaderboard.

phtui detail <product-slug>

Important fields for agents:

• positioning_hint: compact product positioning signal
• target_user_signal: likely audience inferred from category
• monetization_signal: pricing clue when available
• feature_signals: positive/constraint/detail signals to reuse as inspiration
• pros / cons: Product Hunt review signals when available
• maker_comment: launch narrative when available

Plain leaderboard data

Use this when you need raw ranked products without generated inspiration framing.

phtui leaderboard --period daily --limit 10

Agent Response Pattern

When using this skill, answer with a small synthesis instead of dumping JSON:

I found 5 Product Hunt references in AI Agents.

1. Product Name — what it suggests
   - Signal: short market/feature clue
   - Reusable angle: how this could inspire our MVP

Recommended direction:
- Build X first because Y
- Avoid Z until there is stronger evidence

Token Discipline

• Keep CLI --limit small: usually 3-5.
• Fetch detail only for selected slugs.
• Summarize fields; do not paste all JSON.
• Prefer deterministic signals from phtui over asking an LLM to infer from a large raw page.
• Use MCP only when the user's agent workflow specifically requires MCP tools.

Common Pitfalls

1. Using MCP by default. Prefer CLI for low-token inspiration lookup.
2. Requesting too many products. Start with --limit 5; expand only if needed.
3. Treating Product Hunt rank as proof of market demand. Use it as inspiration, not validation.
4. Copying product ideas directly. Extract patterns: audience, workflow, pricing, launch narrative, feature wedge.
5. Forgetting details. Use phtui detail <slug> before making a strong recommendation.

Verification Checklist

• phtui help or go run . help works.
• phtui ideas --category <slug> --limit 3 returns valid JSON.
• At least one selected slug was checked with phtui detail <slug> before final synthesis.
• Final answer includes concrete inspiration patterns, not raw JSON dumps.
• Final answer labels Product Hunt data as inspiration, not conclusive validation.