An automated agent that monitors competitor websites, pricing, and social presence to provide actionable intelligence summaries.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install competitor-watch
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install competitor-watch using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Competitor Watch is a specialized agent skill designed to keep you ahead of the market by automating the tedious task of monitoring rival companies. By leveraging Openclaw Skills, this tool provides a 24/7 watch over websites, product pages, and social channels, ensuring you never miss a strategic move. It uses tiered monitoring to balance depth and frequency, focusing your attention on fierce rivals while maintaining high-level awareness of adjacent market players.
This skill moves beyond simple scraping by implementing intelligent diffing logic. It filters out noise like timestamps and ads to highlight actual product changes, pricing shifts, and new feature announcements. Whether you are a founder or a product lead, this skill transforms raw web data into digestible summaries that help you react strategically to competitor launches before their customers even notice.
To begin using this tool within your Openclaw Skills ecosystem, follow these steps:
scripts/setup.sh
~/.config/competitor-watch/config.json.scripts/add-competitor.sh "CompetitorName" https://example.com --tier=fierce
scripts/check.sh --dry-run
check.sh to your system cron jobs or heartbeat schedule.The skill organizes its data within the ~/.config/competitor-watch directory using a clear hierarchy to ensure data integrity and easy access:
| File/Folder | Description |
|---|---|
config.json |
Main configuration for competitors, tiers, and alert settings. |
data/snapshots/ |
Archive of raw text captures organized by competitor and specific pages. |
change-log.json |
A structured record of every detected significant change across all targets. |
alert-history.json |
Tracking for sent notifications to manage cooldown periods and avoid spam. |
reports/ |
Directory for generated daily or weekly markdown digests. |
Loading
A multi-platform social search engine for AI agents to track mentions, hashtags, and influencers across 1.5 billion indexed posts.

A high-performance search and export tool for Twitter, Instagram, and Reddit data without the need for a developer account.

Perform deep sentiment analysis and social listening across Twitter, Reddit, and Instagram to track brand reputation at scale.

An AI-driven social media research toolkit that enables agents to analyze 1.5 billion posts across Twitter, Instagram, Reddit, and TikTok.

Automatically transform long-form articles, transcripts, and notes into platform-optimized social media threads, posts, and newsletters while maintaining a consistent brand voice.

Trawl is an autonomous SDR agent that sweeps social networks to discover, score, and qualify business leads through semantic search and automated DMs.








































