# Lp Insights

> Generate hypothesis-driven landing-page feedback with simulated personas.

- Skill: `re-sianturi/lp-insights-3` (Agent Skill)
- Install (CLI): `npx skillmds@latest add re-sianturi/lp-insights-3`
- Raw SKILL.md: https://api.skillmd.com/api/skills/re-sianturi/lp-insights-3/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: MIT
- Author: re-sianturi (https://skillmd.com/u/re-sianturi)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/re-sianturi/lp-insights-3

---


# Landing Page Insights (Claude Code Adapter)

Official Claude Code project skill adapter for landing page persona analysis.

## Overview

Analyzes landing pages through simulated personas using project discovery and available Claude Code capabilities.

## Usage

`/lp-insights <url> [traffic_source] [group_persona]`

## Core Loading & Security Procedure

Before execution, the agent must load the following files relative to the skill path:
1. Security guidelines from `core/SECURITY.md`.
2. Workflow rules from `core/WORKFLOW.md`.
3. Persona registry from `core/personas/_registry.md`.
4. Schemas from `core/schemas/persona-result.schema.json` and `core/schemas/final-report.schema.json`.

## Platform Capabilities & Tool Mapping

Claude Code environments vary in available tools depending on user configuration and host permissions.
- **Content Retrieval:** Use host web retrieval tools if present. If unavailable, request the user or environment to provide the rendered page markdown text. Do not invent or assume any tool names.
- **Subagents:** If subagent tools are enabled in the environment, spawn subagents in parallel to execute persona simulations using the canonical prompt contract in `core/prompts/persona-analysis.md`.
- **Sequential Fallback:** If subagent capabilities are disabled or unsupported, run persona simulations sequentially within the primary session context.
- **Relative Path Resolution:** Once installed in a project `.claude/skills/lp-insights/`, persona files are accessed at `personas/` and core files at `core/`.

## Methodology & Disclaimer

- Output declares `evidence_type` as `synthetic_hypothesis`.
- Findings report exact count of affected personas (e.g. 3 of 5 personas affected), never claiming empirical conversion metrics or observed user behavior.