Bluesky Scraper — Profiles, Posts, Followers & More (No Login)

Extract public data from Bluesky through the official AT Protocol AppView — without an account, an app password, or a session token.

Most Bluesky scrapers ask you to hand over your own credentials, which puts your account at risk of rate limits or suspension. This one talks to public.api.bsky.app, the read-only endpoint Bluesky serves to anonymous clients. Nothing to log in with, nothing to leak.

---

11 scraping modes

#	Mode	What you get	Needs
1	profiles	Display name, bio, followers/following/posts counts, verification, avatar, banner, join date, pinned post	handles
2	posts	Author timeline with likes, reposts, replies, quotes, images, video, links, hashtags, mentions	handles
3	followers	Everyone following an account, with full profile metadata	handles
4	following	Everyone an account follows	handles
5	searchPeople	Accounts matching a keyword	searchQuery
6	searchPosts	Posts matching a keyword, with date/language/sort filters	searchQuery
7	postUrls	Specific posts from their bsky.app links (handle → DID resolved automatically)	urls
8	thread	Full conversation: ancestors above the post and a nested reply tree below	urls
9	lists	Public lists and, optionally, every member	handles or urls
10	starterPacks	Starter packs, their join counts, feeds and members	handles or urls
11	feeds	Custom feed generators and the posts they currently serve	handles or urls

Every run also produces a flattened CSV export (OUTPUT.csv in the run's key-value store) alongside the JSON dataset.

---

Quick start

Press Start with the default input — it scrapes the @bsky.app profile and finishes in seconds.

Then adapt one of these:

// Latest 200 posts from an account
{ "mode": "posts", "handles": ["bsky.app"], "maxItems": 200 }

// Followers of several accounts, capped per account
{ "mode": "followers", "handles": ["bsky.app", "jay.bsky.team"],
  "maxItems": 5000, "maxItemsPerSource": 2500 }

// French posts about a topic, most engaged first
{ "mode": "searchPosts", "searchQuery": "intelligence artificielle",
  "searchSort": "top", "searchLanguage": "fr", "maxItems": 500 }

// An entire conversation under one post
{ "mode": "thread",
  "urls": ["https://bsky.app/profile/bsky.app/post/3lkxyzabc123"],
  "threadDepth": 10 }

Handles are flexible: bsky.app, @bsky.app, did:plc:z72i7… and https://bsky.app/profile/bsky.app all work.

---

Input reference

Field	Type	Default	Notes
mode	enum	profiles	One of the 11 modes above
handles	array	["bsky.app"]	Handles, DIDs or profile URLs
searchQuery	string	—	Supports Bluesky search syntax: from:handle, "exact phrase", #hashtag
searchSort	enum	latest	latest or top (mode 6)
searchSince / searchUntil	string	—	YYYY-MM-DD date bounds (mode 6)
searchLanguage	string	—	Two-letter code, e.g. fr (mode 6)
urls	array	—	Post / list / feed / starter-pack links or at:// URIs
maxItems	integer	50	Total cap. 0 means unlimited
maxItemsPerSource	integer	—	Sub-cap so one large account can't eat the budget
includeReplies	boolean	true	Modes 2 and 6
includeReposts	boolean	true	Mode 2
includeListMembers	boolean	true	Modes 9 and 10
threadDepth	integer	6	Reply levels to walk down (mode 8)
threadParentHeight	integer	10	Parent posts to fetch above the target (mode 8)
exportCsv	boolean	true	Also write OUTPUT.csv
proxyConfiguration	object	off	Not needed — the public API has no IP restrictions

---

Output samples

Profile

{
  "type": "profile",
  "did": "did:plc:z72i7hdynmk6r22z27h6tvur",
  "handle": "bsky.app",
  "displayName": "Bluesky",
  "description": "Official Bluesky account.",
  "followersCount": 4512003,
  "followsCount": 42,
  "postsCount": 1337,
  "createdAt": "2023-04-12T04:53:57.057Z",
  "isVerified": true,
  "avatar": "https://cdn.bsky.app/img/avatar/plain/…@jpeg",
  "banner": "https://cdn.bsky.app/img/banner/plain/…@jpeg",
  "profileUrl": "https://bsky.app/profile/bsky.app"
}

Post

{
  "type": "post",
  "uri": "at://did:plc:z72i7…/app.bsky.feed.post/3lkxyzabc123",
  "url": "https://bsky.app/profile/bsky.app/post/3lkxyzabc123",
  "text": "Big update shipping today #bluesky #atproto",
  "createdAt": "2026-05-01T12:00:00.000Z",
  "authorHandle": "bsky.app",
  "likeCount": 5231,
  "repostCount": 940,
  "replyCount": 128,
  "quoteCount": 77,
  "engagementTotal": 6376,
  "hashtags": ["bluesky", "atproto"],
  "mentions": [{ "did": "did:plc:…", "handle": "alice.bsky.social" }],
  "links": ["https://bsky.social/about"],
  "images": [{ "url": "…", "alt": "Screenshot", "width": 1200, "height": 675 }],
  "isReply": false,
  "isRepost": false,
  "isQuotePost": false
}

In thread mode each item carries depth, parentUri and a nested replies array, so you can work with the flat list or rebuild the tree.

---

Pricing

Pay per result — $0.003 per item. You pay only for records actually delivered to your dataset. No monthly platform fee on top, no charge for failed runs.

Job	Items	Cost
One profile	1	$0.003
1,000 posts from an account	1,000	$3.00
10,000 followers	10,000	$30.00

Set maxItems to keep any run inside a known budget.

---

Notes & limits

• Public data only. Anything requiring a session — DMs, notifications, your own feed, blocked/muted lists — is out of scope by design.
• Custom feeds (mode 11). Feed metadata always works anonymously. A few third-party generators refuse unauthenticated reads; when that happens the Actor still returns the generator record and logs a warning instead of failing.
• Deleted or suspended accounts are skipped with a warning rather than aborting the run.
• Rate limits are handled with retry-after-aware backoff. The public API is generous; a proxy is rarely necessary.
• Please respect Bluesky's Terms of Service and applicable data-protection law (GDPR included) when scraping personal data.

---

Local development

pip install -r requirements.txt

# Offline suite: 56 tests, all 11 modes against a mocked AppView
python3 -m pytest tests/ -v

# Live check against the real API (no credentials required)
python3 tests/live_smoke.py

# Run the Actor locally
apify run --input='{"mode":"profiles","handles":["bsky.app"]}'

Project structure

.actor/
  actor.json           Actor manifest
  input_schema.json    Input form definition
  dataset_schema.json  Dataset views in the Apify UI
  pay_per_event.json   Pay-per-event tariff
src/
  __main__.py          Entry point (python -m src)
  main.py              Orchestration, batching, charging, CSV export
  config.py            Input normalisation and validation
  bsky_client.py       AT Protocol client (retries, pagination, DID cache)
  handlers.py          One handler per scraping mode
  transformers.py      Raw AT Proto views -> flat output records
  parsers.py           URL/AT-URI parsing, richtext facets, embeds
  csv_export.py        Nested JSON -> CSV flattening
tests/
  test_offline.py      Unit + mode tests
  test_actor_e2e.py    Full Actor run through the Apify SDK
  live_smoke.py        Real-API smoke test
Dockerfile
requirements.txt