Text expanders don't fully replace the trigger text in Claude Code
context:
i use Espanso to save certain phrases i use. this : para (without space) is used for "Confirm your understanding of my request by paraphrasing it back to me. Do not start any work until I confirm your understanding.". this is so i can easily use phrases i use a lot with various LLMs.
now the slash command serves a similar purpose but the problem is that they can only be triggered at the beginning of writing the message whereas i have a number of prompts that's not really a 'start the message with this' thing, it's more of i write the message 75% of the way and i'm like 'oh yeah it's good to throw that one in'. like : sa for "Use as many subagents as possible to parallelize the work."
ideal situation is that the trigger text gets replaced completely. but this is a minor issue, claude code tends to read my intent perfectly fine even with the remnants of the trigger phrase there.
This issue has 5 comments on GitHub. Read the full discussion on GitHub ↗