sync/downstream.py (9 lines):
	- line 84: # TODO: add PR link to the comment
	- line 101: # TODO: Ensure we have the right set of commits before geting here
	- line 451: # TODO: maybe just get this from GitHub rather than store it
	- line 523: # TODO: Would be nice to do this from mach with a gecko worktree
	- line 638: # TODO: If this commit reverts some closed syncs, then set the metadata
	- line 664: # TODO: Actually check if they landed?
	- line 976: # TODO handle errors here better, perhaps
	- line 1139: # TODO: consider marking the push an error here so that we can't
	- line 1144: # TODO notify relevant people about test expectation changes, stability


sync/upstream.py (7 lines):
	- line 349: # TODO: add label to bug
	- line 422: # TODO - Maybe ignore errors setting the status
	- line 703: # TODO: we should check for this already existing before we process the backout
	- line 716: # TODO: Turn create_syncs into a class
	- line 730: # TODO: Check syncs with pushes that no longer exist on autoland
	- line 816: # TODO: Loading the commits doesn't work in this case, because we depend on the bug
	- line 842: # TODO: Create a new sync with a non-zero seq-id in this case


sync/landing.py (6 lines):
	- line 720: # TODO: this shares a lot of code with downstreaming
	- line 975: # TODO: this shouldn't be mutating here
	- line 999: # TODO: schedule a downstream sync for this pr
	- line 1423: # TODO: should really check if commit is actually part of the sync if there's >1
	- line 1424: # TODO: reopen landing? But that affects the invariant that there is only one
	- line 1446: # TODO: Locking here


sync/sync.py (4 lines):
	- line 98: TODO:  Maybe just store the base branch name in the tag rather than making it
	- line 489: # TODO: Would be good if we could link the repo class to the commit type
	- line 765: # TODO: this object creation is extremely non-atomic :/
	- line 842: # TODO: cancel related try pushes &c.


sync/notify/bugupdate.py (3 lines):
	- line 34: # TODO: not sure the locking here is correct
	- line 83: # TODO: this could make the query over-long; we should probably split
	- line 134: # TODO: handle some more cases here. Notably where the bug is marked as


sync/worktree.py (3 lines):
	- line 32: # TODO: Always cleanup repos where the sync is finished
	- line 184: # TODO: We can get the worktree to only checkout the paths we actually
	- line 226: # TODO: In general the worktree should be on the right branch, but it would


sync/commit.py (2 lines):
	- line 507: # TODO: should bugs be int here
	- line 564: # TODO: Improve the annotations so that this is implied


sync/gh.py (2 lines):
	- line 209: # TODO: Maybe raise an exception here
	- line 558: # TODO: raise the right kind of error here


sync/tc.py (2 lines):
	- line 259: # TODO: not sure if we can avoid tolerating some errors here, but
	- line 328: # TODO: is there a race condition here where dependencies can be


sync/handlers.py (2 lines):
	- line 48: # TODO: maybe want to create a new sync here irrespective of the event
	- line 134: # TODO: other events to check if we can merge a PR


sync/notify/bugs.py (1 line):
	- line 511: # TODO: Ensure that the metadata is added to the meta repo


sync/notify/results.py (1 line):
	- line 445: # TODO: this method should be @mut('sync')


sync/index.py (1 line):
	- line 169: # TODO: @overload could help here


sync/trypush.py (1 line):
	- line 149: # TODO add something useful to the commit message here since that will