feat: add chat name update functionality and improve AI response handling

[Artuomka]

No description provided.

[Copilot]

Pull request overview

Adds automatic chat naming for new AI conversations and tightens AI JSON-output handling to reduce parsing failures.

Changes:

• Generate a chat title on first user message for newly created AI chats and persist it.
• Extend IUserAiChatRepository with updateChatName and implement it in the repository extension.
• Strengthen the AI prompt guidance and post-processing for JSON responses (remove comments / trailing commas).

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
backend/src/entities/ai/use-cases/request-info-from-table-with-ai-v7.use.case.ts	Triggers background chat-title generation for new chats and updates the persisted chat name.
backend/src/entities/ai/ai.service.ts	Updates the LLM prompt to emphasize strict JSON-only output with no comments.
backend/src/entities/ai/ai-conversation-history/user-ai-chat/repository/user-ai-chat-repository.interface.ts	Adds updateChatName to the repository interface.
backend/src/entities/ai/ai-conversation-history/user-ai-chat/repository/user-ai-chat-repository.extension.ts	Implements updateChatName via a TypeORM update query.
backend/src/ai-core/tools/query-validators.ts	Expands AI JSON cleaning to strip comment syntax and trailing commas before parsing.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

generateAndUpdateChatName() already catches and reports errors internally, so the additional .catch(...) here will never run and makes the async fire-and-forget behavior harder to reason about. Either remove the outer .catch and explicitly mark the call as fire-and-forget (e.g., void ...), or remove the internal try/catch and handle errors only at the call site (but avoid double-reporting).

Suggested change

	this.generateAndUpdateChatName(foundUserAiChat.id, user_message).catch((error) => {
	Sentry.captureException(error);
	});
	void this.generateAndUpdateChatName(foundUserAiChat.id, user_message);

[Copilot]

updateChatName is defined without a userId parameter, unlike the other repository methods that scope operations by chatId + userId. To avoid accidental cross-tenant updates, consider changing this API to updateChatName(chatId: string, userId: string, name: string) (or similar) so implementations can enforce ownership at the data layer.

Suggested change

	updateChatName(chatId: string, name: string): Promise<void>;
	updateChatName(chatId: string, userId: string, name: string): Promise<void>;

[Copilot]

This update query scopes only by id. Since chats are user-owned (and other repo methods require userId), this should also include user_id = :userId (and accept userId as a parameter) to prevent updating another user's chat if a UUID is ever reused/leaked/passed incorrectly.

Suggested change

	async updateChatName(chatId: string, name: string): Promise<void> {
	await this.createQueryBuilder()
	.update(UserAiChatEntity)
	.set({ name })
	.where('id = :chatId', { chatId })
	async updateChatName(chatId: string, userId: string, name: string): Promise<void> {
	await this.createQueryBuilder()
	.update(UserAiChatEntity)
	.set({ name })
	.where('id = :chatId AND user_id = :userId', { chatId, userId })

[Copilot]

The block-comment stripping regex (/\/\*[\s\S]*?\*\//g) will also remove /* ... */ sequences that appear inside valid JSON string values, mutating otherwise-correct AI responses (e.g., a description containing those characters). Consider using a comment-tolerant parser (e.g., JSON5) or implementing a small scanner that removes comments only when not inside quoted strings.