Under the hoodTier: how much it should matter to you
1Useful: my rating, 1 to 5
1Signal: worth watching, 1 to 5
SkillsArea: what it touches
Internal ChangesKind: in v2.1.309,
Internal ChangesSection of the release
What
Skills are reusable instructions Claude Code can draw on, and it gives the model a list of them with their descriptions. When that list has to be trimmed to fit a size limit, Claude Code now sends a usage report recording:
how many skills there are
the size limit and the full size before trimming
how the trimming was done
how many skills kept their full description, and how many were cut down
The list itself is the same as before.
Why
Anthropic is measuring how often skill descriptions get cut down to just a name, and nothing about what the model sees changes.
