cf7ed2f66a
Merge pull request #46 from jepler/behavioral-fixes-new-textual
2025-07-24 13:21:10 -05:00
077113632b
core: List available presets in --help
2025-07-24 10:31:43 -05:00
3710a5b93b
Use new textual markdown streaming
2025-07-24 10:30:16 -05:00
eecfa79d29
Merge pull request #44 from jepler/behavioral-fixes-new-textual
2025-05-16 11:36:21 +02:00
5d60927395
Fix app bindings more
...
ctrl-c can now be copy instead of yank (yay!)
2025-04-23 20:19:37 +02:00
633c43502a
fix weird sizing of SubmittableTextArea
...
.. by removing the top & bottom borders. For some reason these appeared
not to be accounted when sizing the widget to its content? not sure.
anyway this improves it.
2025-04-22 20:36:36 +02:00
70cbd2f806
Fix F9 binding of textarea
2025-04-22 20:25:37 +02:00
a8ed8d08d3
Merge pull request #43 from jepler/remove-codeql
...
codeql is not adding value
2025-04-19 20:34:10 +02:00
9aa3a1e910
codeql is not adding value
2025-04-19 20:29:11 +02:00
69566fddbe
Merge pull request #41 from jepler/improve-splat-handling
2024-10-25 08:22:48 -05:00
9340d3551c
Make an API key name of "-" not actually fetch a key
2024-10-23 14:05:18 -05:00
b939bba50f
Improve docs & chap --help
...
Incompatible change: @:FILE presets are now in $HOME/.config/chap/preset
Incompatible change: -S :FILE prompts are now in $HOME/.config/chap/prompt
Incompatible change: -s@FILE is no longer a special case,
use -S FILE instead
2024-10-23 14:05:18 -05:00
993b17845c
Merge pull request #40 from jepler/improve-key-handling
2024-10-23 11:47:20 -05:00
4fc36078a3
Add ability to get keys from a safe password store
...
I use password-store but this can be configured to use anything
that can print a key or secret on stdout.
2024-10-23 11:44:36 -05:00
d38a98ad90
Add UsesKeyMixin & make all key names settable as parameters
2024-10-23 11:43:46 -05:00
8f126d3516
Merge pull request #39 from jepler/misc-updates
2024-10-22 07:57:56 -05:00
06482245b7
Update default model to gpt-4o-mini
...
and give gpt-4o as the common alternative.
2024-10-22 07:56:40 -05:00
64c56e463b
Recommend openai-chatgpt backend over llama-cpp backend
2024-10-22 07:56:40 -05:00
476dfc4b31
Allow setting temperature & top_p in chatgpt backend
2024-10-22 07:56:40 -05:00
e17b06e8d8
Remove a debug print
2024-10-22 07:56:40 -05:00
18616ac187
Merge pull request #38 from jepler/misc-updates
...
Update anthropic back-end & other small changes
2024-06-23 14:48:27 -05:00
0a98049f21
Update get_api to not have any required arguments
2024-06-23 14:46:18 -05:00
415e48496a
update llama backend for llama-3-instruct style
2024-06-23 14:46:18 -05:00
aa1cb90c47
Update anthropic back-end
2024-06-23 14:46:18 -05:00
271792ed8c
Merge pull request #37 from jepler/actions-churn
...
update actions
2024-05-12 21:19:53 -05:00
adec4385ad
one more action bump
2024-05-12 21:17:02 -05:00
5439ded9e8
update actions
2024-05-12 21:12:40 -05:00
ba1adecfbb
Update README.md
2024-05-08 19:46:28 -05:00
2475dc2fcb
Merge pull request #36 from jepler/misc-improvements
...
Various improvements
2024-05-08 19:42:09 -05:00
77297426d6
ruff & mypy fixes
2024-05-08 11:31:31 -05:00
7ea2649640
Commandline improvements
...
* @FILE arguments
* -s:filename searches relative to the configuration path
this makes it easier to create whole preset configurations
2024-05-08 10:10:53 -05:00
90faa3d7a2
click deprecated MultiCommand; the replacement is Group.
2024-05-08 09:01:55 -05:00
f725e01114
openai: Allow specification of API URL
...
This is useful as some providers (including llama.cpp)
have an "openai-compatible API"
2024-05-01 20:40:54 -05:00
79997215ad
Merge pull request #35 from jepler/misc-improvements
...
switch openai default to gpt-4-turbo
2024-04-10 09:56:54 -05:00
08f3fa82d6
fix typing errors
2024-04-10 09:17:27 -05:00
afe22d0fdd
chap ask: add --stdin flag
2024-04-10 08:33:55 -05:00
25dee73c0f
Distinguish an empty system message from an unspecified one
...
.. when deciding whether to use the API's default system message
2024-04-10 08:33:27 -05:00
95a0a29055
Add -@ for reading system prompt from a file
...
`-S @...` did not have good ergonomics for completion.
2024-04-10 08:32:49 -05:00
61af71d0bd
switch openai default to gpt-4-turbo
2024-04-10 08:31:28 -05:00
368275355b
Merge pull request #34 from jepler/mistral
...
Add mistral.ai backend
2024-03-09 20:57:48 -06:00
ddc5214231
Add mistral backend
2024-03-09 20:53:38 -06:00
9eef316a5c
Change up llama.cpp prompting
...
this gives good results on mixtral.
2024-03-09 20:51:16 -06:00
fac2dfbdc3
Merge pull request #33 from jepler/anthropic-claude
...
Add an anthropic back-end
2024-03-08 14:26:56 -06:00
d63d8e6fe2
Add an anthropic back-end
2024-03-08 14:19:31 -06:00
2995b1e1aa
Merge pull request #31 from jepler/reduce-code-block-jank
...
Show code blocks at full height
2023-12-12 21:50:08 -06:00
59d9e68119
Show code blocks at full height
...
...well, for any reasonable height anyway.
When scrollbars appear the refreshing output tends to get janky, so
just allow the code region (called a fence in markdown, apparently)
to be up to 9999 lines tall.
2023-12-12 21:38:23 -06:00
9f6ace394a
Merge pull request #30 from jepler/multiline2
...
Switch chap tui to a multiline text field
2023-12-12 17:05:05 -06:00
86059a5d85
re-work backend parameters again to work on 3.9 and 3.11
2023-12-12 17:01:41 -06:00
877246ac28
document chap tui a bit more in the README
2023-12-12 16:08:54 -06:00
a42a349bb7
fix typing of main function
2023-12-12 16:03:08 -06:00
