v0.2.15

Changelog

• 1183860 Better buffer insertion
• e265d50 Optimize processing shellbuffers with lots of mid-buffer insertions
• 129ced2 Use string instead of array for completion requests
• fbb59f0 Merge pull request #34 from schnerd/patch-1
• 70c0157 Fix env var in command description

v0.2.14

Changelog

• 815d6de Fix panic on no autosuggest encoder found
• 1016527 Improve light mode colors and fix goal mode color bug
• d6b602c Revert default encoder
• 525e17a Merge pull request #30 from sash20m/gpt-4o
• 237b0d4 Added GPT-4o

v0.2.13

Changelog

• 2a69b7d Upgrade go and deps
• 23a494d Add max prompt tokens shell argument
• fd567e5 Handle missing models in tiktoken better

v0.2.12

Changelog

• 7d87f25 Fix broken prompting introduced in v0.2.11 and set tokentimeout in more places

v0.2.11

Changelog

• d7573b9 Add config for answer length, set to 2048, give more succinctness instructions
• 744e8ce Upgrade dependencies
• 0892460 Add new models and default to gpt-4-turbo
• 77a4d60 Set time between token timeout
• 0eafb0b Update default model and max history block tokens
• 6992df1 Move pretty ansi code to its own area

v0.2.10

Changelog

• 66be1ff Look in OPENAI_API_KEY env var for credential
• 5537314 Link butterfish.nvim

v0.2.9

Changelog

• 021134a Reset styled code writer in between prompts

v0.2.8

Changelog

• 4b7319a Fixing markdown block formatting

v0.2.7

Changelog

• 722e0d0 Fix bugs with styling when markdown blocks were indented

v0.2.6

Changelog

• d6d6e34 Tweak answer colors
• bd9496c Add special style for inline markdown (single backticks)