A Bluesky list is good for private curation. A starter pack is good when you want other people to join the same slice of Bluesky.

I do not treat them as the same thing.

Use TheBlue.social's Bluesky list and starter pack converter when you have a list or starter pack that is almost right, but you want to review the accounts before creating the other version.

Start with the job of each format

The official Bluesky user lists documentation says every list has a purpose.

For regular lists, the useful one here is app.bsky.graph.defs#curatelist: a list of users for feeds or thread gates. A custom list can be a reading list, a research queue, a relationship tracker, or a small account set you keep for yourself.

Bluesky's official starter packs post describes starter packs as personalized invites that can recommend custom feeds and users. The same post says a starter pack can include up to 150 people and up to 3 custom feeds.

So I use the formats differently:

  • a custom list is my workspace
  • a starter pack is my recommendation

That distinction keeps the workflow simple.

Keep a custom list while you are still researching

I start with a custom list when the account set is not ready for other people.

I usually stay in list mode while I am still figuring out:

  • which accounts are active
  • which accounts are duplicates or mirrors
  • which accounts are useful but too noisy
  • which accounts are friends of the topic, not core members
  • whether the list needs subtopics
  • whether I want to keep the list private in practice, even if the Bluesky list itself is public

A list lets me work slowly.

I can add accounts as I find them, remove bad fits, and use the list as a feed while I learn what the community posts every week. If I turn it into a starter pack too early, I make other people inherit my unfinished research.

For example, a "local Singapore tech" list might start with founders, engineers, community organizers, event accounts, recruiters, investors, and a few company accounts. The mix helps during research. In a starter pack, I would split it into smaller account sets.

The list is where I sort that out.

Convert when the list has a clear promise

I turn a list into a starter pack when I can explain it in one sentence.

Good examples:

  • "People posting useful climate science."
  • "Indie game developers who share work in progress."
  • "Local journalists covering Singapore and Southeast Asia."
  • "Bluesky accounts for web developers who write about shipping products."

Weak examples:

  • "Interesting people."
  • "Accounts I might want later."
  • "People from a thread I liked."
  • "Everyone in this broad community."

The difference matters because starter packs are onboarding objects. Someone opens one because they want a fast answer to "Who should I follow for this topic?"

If the pack promise is vague, the follow decision gets harder.

Review before converting

I review the list before I create the starter pack.

My pass is simple:

  1. Remove inactive accounts.
  2. Remove accounts that post mostly outside the topic.
  3. Remove accounts I added only for monitoring.
  4. Remove duplicates, mirrors, and company accounts unless they belong.
  5. Keep the accounts a newcomer would actually benefit from following.

This is why I like converting through a tool instead of copying everything by hand.

The converter lets me paste the list, search the members, remove bad fits, then create the starter pack from the remaining accounts. I get the review step without rebuilding the account set one handle at a time.

I also keep the source link in the description:

Source: https://bsky.app/profile/example.com/lists/3abc123
Converted from a custom list after removing inactive and off-topic accounts.

Future me needs that note.

Check the pack like a new follower

The review pass changes when I imagine the pack from a new user's view.

I am no longer asking, "Do I want this account somewhere in my research system?" I am asking, "Would this account make sense to someone who just clicked a starter pack title and wants a useful first follow set?"

That question removes a lot of borderline accounts.

An account can be worth monitoring and still be a poor recommendation. It may post too rarely, use the topic only during one event, mostly repost other people, or need context that a new follower will not have.

I check:

  • the display name and handle
  • the bio
  • recent posts
  • whether the account still posts about the topic
  • whether the account mostly replies or posts original material
  • whether the account would surprise someone who followed the pack title

I am trying to avoid handing people a noisy first experience.

If I am unsure, I leave the account in the list and keep it out of the starter pack. Lists are allowed to be messy. Starter packs should earn more trust.

Watch the 150-account limit

The official Bluesky starter-pack post says a starter pack can recommend up to 150 people.

The limit helps because it forces a decision.

If my source list has 500 accounts, I do not create one giant starter pack in spirit and pretend it is focused. I split it.

For a broad "software people" list, the better starter packs might be:

  • indie software builders
  • web platform people
  • game developers
  • open source maintainers
  • local tech community accounts

Each pack gets a clearer promise. Each one is easier to follow.

The list can stay broad. The starter pack should be tighter.

This also keeps maintenance realistic.

A 150-account pack is possible, but I rarely want to review 150 people in one sitting. A smaller pack is easier to revisit after a conference, product launch, spam wave, or community split.

When a pack has a narrow promise, I can spot drift faster. If the promise is "local journalists covering Singapore and Southeast Asia," a brand account or inactive conference account stands out. If the promise is "interesting people," everything fits and nothing fits.

The source list can keep the long tail. The starter pack should be the part I am comfortable recommending today. Smaller packs also get cleaned more often.

Convert the other direction when you need a working list

Sometimes I go the other way.

A starter pack is good for onboarding, but it is not always the format I want for ongoing work. If I find a starter pack that looks useful, I may convert it into a custom list first.

That gives me a working list I can:

  • read as a focused feed
  • edit over time
  • use as a review queue
  • split into smaller lists
  • copy later into a starter pack I maintain

This helps when someone else's starter pack is close to what I need and still needs editing for my feed.

I do not want to follow everyone immediately. I want to inspect the accounts, remove the ones that do not fit, then decide whether the result should be a list, a starter pack, or just a temporary research set.

Keep the source trail

When I convert a starter pack into a list, I still keep a source note.

Source notes feel fussy until I return to the list a month later. Without a note, I have to remember whether I built the list from scratch, copied it from a starter pack, or merged it from three different sources.

The source trail also makes future edits easier. If the original starter pack is still maintained, I can compare my list with the newer version instead of guessing what changed.

For a working list, the note can be short:

Source starter pack: https://bsky.app/starter-pack/example.com/3abc123
Converted to a custom list for review before following.

I only need enough context to make the next review less annoying.

My default workflow

Here is the workflow I use:

  1. Build or find a Bluesky custom list.
  2. Use it for a while as a focused feed.
  3. Remove inactive, off-topic, duplicate, or monitoring-only accounts.
  4. Decide the one-sentence promise.
  5. Keep the list broad if I still need it for research.
  6. Convert it to a starter pack only when it helps someone else follow the topic.
  7. Keep the source URL and edit note in the description.
  8. Review the starter pack later, especially after events or community shifts.

The list is where I think. The starter pack is what I share.

Last updated: June 13, 2026