ScrapeAtlas CLI
Official ScrapeAtlas terminal client: an agent or script pulls public data from 49 platforms with one command and gets plain JSON
Broad access
Why this level
- The customer key sits in a config file as plain text, and an agent holding it spends paid credits and can drain the balance with a burst of requests
- Requests and responses pass through the vendor's server, and answers contain data about accounts and people that you need the right to collect, within each platform's terms
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/scrapeatlas-cliDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- The customer key sits in a config file as plain text, and an agent holding it spends paid credits and can drain the balance with a burst of requests
- Requests and responses pass through the vendor's server, and answers contain data about accounts and people that you need the right to collect, within each platform's terms
This is third-party code. Review the repository files before installing.
What it does
ScrapeAtlas CLI calls the ScrapeAtlas API, which returns structured public data from social networks, video services, forums, marketplaces and ad libraries. A command such as scrapeatlas bluesky profile --handle bsky.app sends one request to the api.scrapeatlas.com gateway and prints the full JSON to stdout, so jq or a script can parse it right away. The list of operations ships inside the package. scrapeatlas list shows the platforms, and scrapeatlas list reddit --json prints operations with their IDs and request schemas, so an agent can find the right one on its own. Parameters go in as flags, with kebab-case names accepted alongside the original ones. Arrays and nested objects go through --input, including from a file such as @request.json. The next page is requested by hand: pass the cursor from the response to --cursor. Every call makes a single request, with no retries, redirects or media downloads. A request error ends the command with code 1 and keeps the response body, so an empty result is distinguishable from a failure. A usable but partial answer exits with 0, and you then check the status and coverage fields. Output can be written to a new file with --output or wrapped in an envelope with the HTTP code using --envelope.
Who it is for. For developers, analysts and marketers who need public platform data for research, brand monitoring and AI workflows and want to call the API from a terminal, scripts and agents such as Claude Code, Codex or Cursor. It needs Node.js 22 or newer and a ScrapeAtlas customer key.
Good fit when
- You need public posts and comments on a topic from Reddit or Hacker News, parsed with jq
- An agent has to read a product card, reviews and seller offers on a listed marketplace and save the answer to a file
- You study competitors' public ad creatives in the Meta and LinkedIn ad libraries
- Before integrating through REST or n8n you want to see in a terminal which fields a given operation returns
- Your brand's public profiles and videos are checked on a schedule in CI, with the key kept in an environment variable
Not a fit when
- You need private data: direct messages, locked accounts, content behind a login. The service terms forbid requesting it
- You need an MCP server. The package does not run one; ScrapeAtlas offers MCP separately, marked as a preview on its site
- You need downloaded images and video or automatic walking through every page: the CLI returns JSON only and moves one page per call
- You have no right to collect this data, for example about people when you have not established a lawful basis for processing
- You need free unlimited collection: every successful request spends credits, and only key checks and the command list are free
Example request
Use scrapeatlas to take 20 recent posts from a public subreddit about MCP, save the response to posts.json and list the discussion topicsLimitations
The service is paid and foreign: the contract is with a company in the emirate of Sharjah, UAE, and credit packs are sold in US dollars. A new account gets 100 free credits, credits do not expire and there are no subscriptions. An account accepts up to 120 requests per minute and four concurrent ones.
Requests and responses go through the vendor's gateway. ScrapeAtlas does not guarantee its sources: an answer can be partial or have the challenged status, and completeness and freshness depend on the source platform. The service is independent and not affiliated with the platforms.
The tool collects data from other people's platforms, so apply it only to data you have the right to collect and follow each platform's terms. The service terms forbid bypassing logins, paywalls and access restrictions, and public availability does not lift personal data law requirements. The package is distributed through npm under the MIT license and carries no link to a source repository.
How to disable. Remove the package with npm uninstall -g @scrapeatlas/cli. scrapeatlas auth logout erases the saved key from ~/.config/scrapeatlas/config.json but does not revoke it. Revoke the key on the keys page of the ScrapeAtlas dashboard, and remove the SCRAPEATLAS_API_KEY variable from your environment and CI.
FAQ
How do I install ScrapeAtlas CLI and connect it to Claude Code or Codex?
Install Node.js 22 or newer, then run npm install -g @scrapeatlas/cli and scrapeatlas auth login: a hidden prompt verifies the key through the gateway and saves it. The agent needs no separate plugin and runs scrapeatlas commands in the shell. For CI and agents the SCRAPEATLAS_API_KEY variable is safer, because the --api-key flag stays in shell history.
How much does ScrapeAtlas cost and is there a free tier?
An account gets 100 free credits with no card. According to the site as of October 28, 2026, packs are one-time purchases: 10,000 credits for 19 dollars, 50,000 for 89, 200,000 for 249. A successful CLI call costs one credit, and key checks and account views are free.
Can I see the command list without a key?
Yes. npx @scrapeatlas/cli list prints platforms and operations without installing or using a key. The key is needed only for data requests.
What does the CLI do when a platform returns a challenge or an error?
The command exits with code 1 and the full response body stays on stdout. The CLI makes no retries, redirects or workarounds: you or your script decide whether to send a new request.
How is ScrapeAtlas CLI different from Apify MCP Server and Firecrawl CLI?
ScrapeAtlas offers a fixed set of ready operations for 49 platforms in one response schema. Apify runs scrapers from its own store, and Firecrawl reads arbitrary websites. If you need Reddit, YouTube or Amazon in the same JSON shape, ScrapeAtlas fits.
Discussion
No questions or reviews yet.
Related
CLI toolsA data visualization skill for agents: turns data into polished interactive HTML charts and full reports with one visual language
A CLI and 13 skills for Claude Code: the Power BI data model, DAX measures, row-level security and PBIR report pages straight from chat
A skill that teaches an agent to search tables, trace lineage and read the glossary in your Marmot catalog via CLI, REST and MCP.
Customer feedback analysis on your own server: personal data is stripped before the model, and an MCP server and skill give the agent quotes from reviews
A community collection of Apify skills: ready routes to Actors for leads, jobs, prices, ads, Reddit and Yandex search
High-Stakes Analytics Lab
High-Stakes Analytics & Decision Lab
An evidence-gated analytics skill for high-stakes decisions: profiles data, picks a method the evidence supports, can answer with a data request
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| ScrapeAtlas CLI (this entry) | yes | |||
| Apify MCP Server | yes | OAuth | 13,841 | runs arbitrary scrapers from the Apify store over MCP, while this one offers ready per-platform operations in one response schema |
| Firecrawl CLI | no | 674 | reads arbitrary websites, while ScrapeAtlas returns structured data from a fixed set of platforms |
README in short
The CLI documentation walks from installation and login to finding operations, passing parameters and handling output. Key storage, key source priority and exit codes get their own sections. It ends with the MIT license and the support address.