add option to always send full page context

allow models to have a zero token limit

This allows for configuring completely unknown models. For these models
no token limit is known and we will simply do not apply any. Instead we
trust that the

allow models to have a zero token limit

This allows for configuring completely unknown models. For these models
no token limit is known and we will simply do not apply any. Instead we
trust that the model will be either large enough to handle our input or
at least throw useful error messages.

show more ...

new UI with option to chat about the current page

added configurable custom prompt

This makes it easier to add to the existing internal prompt instead of
completely overwriting it using the localization mechanism.

auto style fixes

ask the rephrased question only if it has more context

Otherwise it's usually better to let the LLM reevaluate the context.

mechanisms to override things on command line

This should help with debugging/evaluating

automatic stylefixes

prefer prompted user messages over system prompts

This seems to be better supported across different model providers and
seems to influence the model in a stronger way.

Prompt files have been renam

prefer prompted user messages over system prompts

This seems to be better supported across different model providers and
seems to influence the model in a stronger way.

Prompt files have been renamed to avoid having them show up in the
DokuWiki translate interface. Multilingual models are fine with
prompting in english only.

show more ...

various refactoring and introduction of a simulate command

The new command makes it easier to run the same chat questions against
multiple models and compare the results in a spreadsheet

separate the rephrasing model from the chat model

Rephrasing can be done with faster, simpler models as there is not much
reasoning needed.

correctly use storage setting

move model configuration into json files

This removes the use of individual classes for each model in favor of
more general client classes that only get a model name passed. The meta
info about pric

move model configuration into json files

This removes the use of individual classes for each model in favor of
more general client classes that only get a model name passed. The meta
info about price and token limits is configured in a json file

show more ...

more refactoring on chat and embed model support

* differentiate between input and output tokens
* make use of much larger input contexts

first tries with Anthropic

Use interfaces for Chat and Embedding classes

This way it's easier to have a base OpenAI class. This also moves much
of the statistics and http handling into the base class making model
implementati

Use interfaces for Chat and Embedding classes

This way it's easier to have a base OpenAI class. This also moves much
of the statistics and http handling into the base class making model
implementations even leaner

show more ...

First start on refactoring the class hierarchy

This splits embedding models from chat completion models.

tell the model to fall back to current language if it's unsure

The model defaulted to spanish if no sources were found

record the times of embed and maintenance runs

This makes it easier to debug when something with the cronjob goes
wrong. Currently the data is only exposed in the cli info command.

We might want to

record the times of embed and maintenance runs

This makes it easier to debug when something with the cronjob goes
wrong. Currently the data is only exposed in the cli info command.

We might want to use it somewhere in the UI to warn about outdated data

show more ...

Added Qdrant Storage

fixed overlong lines

�� Automatic code style fixes

Added ChromaDB vector storage support

codesniffer cleanups

�� Automatic code style fixes