Module chatbot_configurations

Structs§

ReasoningEffortLevel
ToolCategory: A category of chatbot tools a configuration can choose to offer the LLM. Independent of the chatbot crate’s per-tool ToolPermission check: a category answers “does this chatbot offer this kind of tool”, not “may this caller use it”.
ToolCategoryGroup: The UI/authorization grouping a ToolCategory belongs to. Derived from the leaf, never stored: the database and the wire format only ever carry ToolCategory values.
VerbosityLevel

MIN_MAX_OUTPUT_TOKENS 🔒: Minimum max_output_tokens allowed for a configuration. Too small a budget cannot produce a usable response — and with reasoning models the hidden reasoning tokens are spent from the same budget, so the floor needs to leave room for the actual answer either way.

azure_search_allowed 🔒: Whether the configuration may use Azure search, which needs a course: the search index it queries is built per course, so a configuration without one has nothing to search.
delete
edit
get_all_chatbots
get_by_id
get_enabled_nondefault_for_course
get_for_azure_search_maintenance
get_for_course
insert
normalized_tool_categories: Dedupes and sorts into ToolCategory::ALL order, so the stored array is canonical and two equal sets (e.g. compared by the edit_chatbot controller) compare equal regardless of the order the caller submitted them in.
remove_default_chatbot_from_course
set_default_chatbot_for_course
validate_max_output_tokens 🔒: Rejects configurations whose max_output_tokens is too small to produce a usable response.