Bluesky Scraper - Posts, Hashtags & Engagement

Extract posts from Bluesky, the AT Protocol network. Search by keyword or pull a specific account's feed, and export each post with its engagement counts, hashtags and media links.

What you get per post

Field	Example
text	the post content
authorHandle, authorDisplayName, authorDid	handle, shown name and the stable AT Protocol identifier
likeCount, repostCount, replyCount	engagement
hashtags	["#dev", "#opensource"]
hasImages, imageUrls	attached media
langs	declared post languages
indexedAt	when the network indexed it
postUri, postUrl	AT Protocol URI and a normal web link

Search filters

• searchQuery - keyword, phrase or hashtag to search for
• actorHandle - a specific account whose posts to collect, e.g. bsky.app
• language - restrict to a post language
• sort - latest or top results
• maxResults - how many posts to save

Authentication

Bluesky's search endpoints require a logged-in session. Supply your own account:

• identifier - your Bluesky handle or email
• appPassword - an app password, created in Bluesky under Settings, Privacy and Security, App Passwords

Use an app password, never your main password. App passwords can be revoked individually and cannot change your account settings.

Example input

{
  "searchQuery": "#opensource",
  "sort": "latest",
  "maxResults": 200,
  "identifier": "yourhandle.bsky.social",
  "appPassword": "xxxx-xxxx-xxxx-xxxx"
}

Use cases

• Brand and product monitoring - track mentions of your name or category as they happen
• Social listening - measure how a topic is discussed, and by whom, on a network that is still open
• Influencer discovery - rank accounts in a niche by engagement rather than follower count
• Trend research - follow hashtag volume over time
• Competitive intelligence - watch what competitors post and how it lands
• Dataset building - authorDid is a stable identifier, so longitudinal datasets survive handle changes

Why Bluesky

The AT Protocol is genuinely open: posts, likes and reposts are public records rather than platform-internal metrics, and accounts carry a permanent DID independent of their handle. For research that has to survive renames and platform churn, that is a real advantage over closed networks.

Notes

• Engagement counts are snapshots at collection time; re-running the actor on the same query shows how a post accumulates.
• authorDid is permanent, authorHandle is not. Key your data on the DID.
• The app password is used only to create a session with Bluesky and is never written to the dataset.