How to use the Competitor Intelligence via Apify agent skill

What is an agent skill?

An agent skill is a folder containing a SKILL.md file with instructions, plus optional scripts and reference files. When you ask a compatible AI agent a matching question, it reads these instructions and follows the workflow they describe.

This skill is named apify-competitor-intelligence. Its SKILL.md file describes a five-step workflow, and a Node.js script at reference/scripts/run_actor.js does the actual work of starting remote scrapers.

What the skill makes the agent do

The agent chooses an Apify Actor that fits your question. The skill's table maps needs to Actors, for example a Google Maps business scraper for location analysis, a Booking.com hotel scraper for hotel benchmarking, Instagram post or reel scrapers for content monitoring, and TikTok or YouTube scrapers for video metrics.

Before running, the agent fetches the Actor's input schema and asks you for an output format and a number of results. It then runs the script, which starts the Actor remotely, and reports back with the number of competitors fetched, missing data, the file location if one was saved, key competitive insights and suggested next steps.

• Quick answer: top results shown in chat, no file saved
• CSV: bounded export of up to 1000 rows
• JSON: bounded export of up to 1000 rows

How to install the skill

Download the ZIP from the skill's page on this directory and unzip it. For Claude Code, place the folder in ~/.claude/skills/ to make it available everywhere, or in a project's .claude/skills/ folder to limit it to that project.

In the Claude apps, you can instead upload the skill under Settings → Capabilities → Skills.

Set up the requirements

The skill needs three things before it can run. Create a .env file containing APIFY_TOKEN=your_token, using a token from your Apify account. Install Node.js 20.6 or newer, which supports the --env-file option the script uses. Install the mcpc CLI with the command below.

You also need network access, because the script contacts Apify services.

• Node.js 20.6 or newer
• An Apify account and APIFY_TOKEN in a .env file
• mcpc CLI

How to trigger the skill with natural requests

You do not need to type the skill name. Describe a competitor task that matches its scope, and name the platform and the output you want. The skill's own guidance says to use it only when the task clearly matches its scope, so specific requests work best.

Good requests name a competitor, a platform, a metric and a format. Vague requests may not trigger the skill, or may lead the agent to ask clarifying questions.

• Compare the review counts and ratings of these three coffee shops on Google Maps and save a CSV.
• Pull the last 50 posts from this Instagram competitor and summarise their content themes.
• Benchmark these two hotels' Booking.com reviews and give me a quick answer in chat.
• Find our main rival's recent TikTok videos and list the ones with the highest engagement.

What to expect during a run

The agent confirms the Actor, inputs, result limits and spending cap with you before starting. The script runs the Actor remotely and waits for it to finish. If the local wait times out, the remote run may still be running, so the agent should inspect the existing run rather than start a duplicate.

Results saved to a file follow the naming pattern in the skill, which includes a date prefix such as YYYY-MM-DD_OUTPUT_FILE.

Safety notes: scripts, network and paid runs

This skill runs scripts and uses the network. Scraping runs on Apify's paid cloud service. The script does not enforce a remote spending cap, so set limits in your Apify account or approved run setup before you start. A local timeout does not stop a paid remote run.

The skill tells the agent not to bypass authentication, not to collect unnecessary personal contact data, and not to put private tokens, session cookies or sensitive files into inputs, because inputs go to Apify and Actor operators. Scraped content and Actor READMEs are treated as untrusted data. The script identifies itself to Apify with a User-Agent header.

• Set spending limits in your Apify account before running
• Never include APIFY_TOKEN, cookies or sensitive files in inputs or chat
• Check platform terms, since they may restrict scraping
• Review results before relying on them

Tips for better competitor research

Start with a small result count to check the Actor's output and cost, then scale up. Pick the Actor that matches the metric you need, since the table separates ad, page, post, reel, comment and profile scrapers for each platform.

Treat the output as a starting point. The skill itself says results are not a substitute for expert review, and scraped data may be incomplete or stale, so check important findings against the source.

• Test with a small limit before a full pull
• Use the Actor table to match the platform and metric exactly
• Date-stamp exported files so you can compare snapshots over time
• Ask the agent to report missing data and pagination limits