Support syncing from local version-controlled CSV

[patcon]

Builds on #2

CSV sync would allow clearer support of multiple lists, and also allow for more easily supporting sync from pre-existing spreadsheet resources.

Proposed format:

bsky_username	bsky_xxxxxx	bsky_yyyyyyy
patcon.bsky.social	x
jaseemabid.bsky.social		x
foo.bsky.social	x
bar.bsky.social	x

where xxxxx / yyyy are values that somehow link to the list (more on this below)

Pros:

• column names are explicit, and could be added to existing spreadsheets
• can add extra columns for whatever metadata someone wants to track in the process of adding ppl (e.g., cohort year, role, etc depending on group using this)

Cons:

• a little less simple

Open Questions:

1. is xxxxx/yyyyyy the direct starter pack / list URI? or is it a human-friendly word that's joined in another config somewhere, e.g., an envvar?

[jaseemabid]

Currently the code syncs the same set of accounts to 2 lists. After poking around the bsky API a bit, it looks to me that starter packs are just another list with some additional metadata, so I might be able to simplify the code there a bit as well.

Could do something along the lines of sync.py --mapping bsky_xxxxxx=at://one --mapping bsky_yyyyyyy=at://two or something similar to be explicit.