Second phase of the AI feature. Users can now open a slide-in chat drawer from the SiteHeader and ask natural-language questions about their company's invoices, customers, payments, and expenses. The LLM invokes pre-defined read-only tool functions (scoped to the current company at execute time) to fetch data and synthesize answers.
**Database** — new ai_conversations and ai_messages tables. Messages are stored in OpenAI's chat format so AiAssistantService serializes a conversation into an API request with zero translation. Columns: role, content, tool_call_id, tool_calls JSON, model, tokens_in, tokens_out. Conversations are scoped (company_id, user_id) — one user's chats are invisible to everyone else, even inside the same company. Foreign-key cascade deletes.
**Tool infrastructure** — AiTool abstract base + AiToolRegistry singleton (registered in a new AiServiceProvider). The base class enforces the security rule: every tool's execute() receives companyId and userId as injected parameters; tools' JSON schemas NEVER include a company_id field. An LLM physically cannot pass a company_id and escape tenancy. Modules can register their own tools by resolving the registry from their own ServiceProvider::boot().
**Nine built-in tools**: search_invoices, get_invoice, search_customers, get_customer, list_recent_payments, list_overdue_invoices, get_company_stats (aggregates for named periods), search_items, list_expense_categories. All read-only; no mutations. Each returns JSON-encodable data the LLM can parse.
**AiAssistantService orchestration loop** — the heart of Phase 2. Flow: persist user message → build payload from system prompt + recent history (40-message window) + new user message → call driver.chatCompletion with tools → if tool_calls, execute each one via the registry (with injected scope), persist tool result, loop → if plain text, persist and return. Hard cap at 5 iterations to prevent runaway LLMs. System prompt pins the assistant to this company's data and forbids mutation.
**Controllers + policy + rate limit** — POST /api/v1/ai/chat runs the orchestration loop. GET/PATCH/DELETE /api/v1/ai/conversations for CRUD. AiConversationPolicy enforces user_id+company_id match on every action. A new 'ai' RateLimiter in RouteServiceProvider throttles to 30 req/min per (user, company). New 'use ai' Gate defined in AppServiceProvider returns true for any authenticated user — the per-company kill-switch still goes through AiConfigurationService::resolveForCompany.
**Frontend** — new features/company/ai/ folder with a Pinia store (ai-chat.store.ts) holding drawer state, current conversation, messages, and loading flags. AiChatDrawer.vue is a slide-in panel teleported to <body>, mounted globally in CompanyLayout.vue when bootstrap reports ai.enabled && ai.chat_enabled. Sub-components: AiChatMessage (user bubbles vs assistant bubbles), AiChatMessageInput (Enter submits, Shift+Enter newline), AiChatConversationList (sidebar with 'new chat' button, rename, delete). A SparklesIcon button in SiteHeader toggles the drawer.
**Driver test double** — tests use a ScriptedAiDriver registered via AiDriverFactory::register('scripted', ...) that returns pre-queued AiChatResponse objects. Feature tests cover: happy path (new conversation + message persistence), tool-call loop (multi-round-trip with search_invoices), runaway-loop cap, driver-throws path, ai_enabled=NO rejection, chat role disabled rejection, per-user conversation visibility, cross-user policy enforcement, cascade delete.
388 tests pass (was 372, +16 new). Pint clean. npm run build clean. Phase 3 (WYSIWYG text generation popup) is the remaining follow-up.
* Convert string references to `::class`
PHP 5.5.9 adds the new static `class` property which provides the fully qualified class name. This is preferred over using strings for class names since the `class` property references are checked by PHP.
* Use Faker methods
Accessing Faker properties was deprecated in Faker 1.14.
* Convert route options to fluent methods
Laravel 8 adopts the tuple syntax for controller actions. Since the old options array is incompatible with this syntax, Shift converted them to use modern, fluent methods.
* Adopt class based routes
* Remove default `app` files
* Shift core files
* Streamline config files
* Set new `ENV` variables
* Default new `bootstrap/app.php`
* Re-register HTTP middleware
* Consolidate service providers
* Re-register service providers
* Re-register routes
* Re-register scheduled commands
* Bump Composer dependencies
* Use `<env>` tags for configuration
`<env>` tags have a lower precedence than system environment variables making it easier to overwrite PHPUnit configuration values in additional environments, such a CI.
Review this blog post for more details on configuration precedence when testing Laravel: https://jasonmccreary.me/articles/laravel-testing-configuration-precedence/
* Adopt anonymous migrations
* Rename `password_resets` table
* Convert `$casts` property to method
* Adopt Laravel type hints
* Mark base controller as `abstract`
* Remove `CreatesApplication` testing trait
* Shift cleanup
* Fix shift first issues
* Updating Rules for laravel 11, sanctum config and pint
* Fix Carbon issue on dashboard
* Temporary fix for tests while migration is issue fixed on laravel side
* Carbon needs numerical values, not strings
* Minimum php version
* Fix domain installation step not fetching the correct company_id
* Fix Role Policy wasn't properly registered
---------
* Create PHP CS Fixer config and add to CI workflow
* Run php cs fixer on project
* Add newline at end of file
* Update to use PHP CS Fixer v3
* Run v3 config on project
* Run seperate config in CI
