mirror of
https://github.com/InvoiceShelf/InvoiceShelf.git
synced 2026-07-16 22:05:20 +00:00
e861fc1fc1
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.
92 lines
2.5 KiB
PHP
92 lines
2.5 KiB
PHP
<?php
|
|
|
|
namespace App\Services\Ai;
|
|
|
|
use App\Services\Ai\Tools\AiTool;
|
|
use InvalidArgumentException;
|
|
|
|
/**
|
|
* In-memory registry of AiTool instances.
|
|
*
|
|
* Register tools from a service provider (see `App\Providers\AiServiceProvider`)
|
|
* at app boot; the AiAssistantService reads `schemas()` to populate the LLM's
|
|
* tool-calling payload and calls `execute()` when the model returns a tool_call.
|
|
*
|
|
* The registry is intentionally a singleton: tools themselves are stateless
|
|
* dispatchers, so one instance shared across the request is safe. Modules can
|
|
* register their own tools by resolving this service and calling `register()`
|
|
* from their own ServiceProvider::boot().
|
|
*
|
|
* $this->app->resolving(AiToolRegistry::class, function (AiToolRegistry $registry) {
|
|
* $registry->register(new MyCustomTool);
|
|
* });
|
|
*/
|
|
class AiToolRegistry
|
|
{
|
|
/**
|
|
* @var array<string, AiTool>
|
|
*/
|
|
protected array $tools = [];
|
|
|
|
public function register(AiTool $tool): void
|
|
{
|
|
$this->tools[$tool->name()] = $tool;
|
|
}
|
|
|
|
/**
|
|
* @return array<string, AiTool>
|
|
*/
|
|
public function all(): array
|
|
{
|
|
return $this->tools;
|
|
}
|
|
|
|
public function get(string $name): ?AiTool
|
|
{
|
|
return $this->tools[$name] ?? null;
|
|
}
|
|
|
|
/**
|
|
* Export all registered tools as the `tools` array for an OpenAI-style chat request.
|
|
*
|
|
* @return array<int, array<string, mixed>>
|
|
*/
|
|
public function schemas(): array
|
|
{
|
|
return array_values(array_map(
|
|
fn (AiTool $tool): array => $tool->toOpenAiToolSchema(),
|
|
$this->tools,
|
|
));
|
|
}
|
|
|
|
/**
|
|
* Execute a tool by name, injecting company + user scope from the caller's session.
|
|
*
|
|
* The AiAssistantService is the only place this should be called from — that's
|
|
* how we guarantee the `$companyId` and `$userId` arguments are session-authoritative
|
|
* and never influenced by LLM output.
|
|
*
|
|
* @param array<string, mixed> $arguments
|
|
*
|
|
* @throws InvalidArgumentException When the tool name is not registered.
|
|
*/
|
|
public function execute(string $name, array $arguments, int $companyId, int $userId): mixed
|
|
{
|
|
$tool = $this->get($name);
|
|
|
|
if ($tool === null) {
|
|
throw new InvalidArgumentException("Unknown AI tool: {$name}");
|
|
}
|
|
|
|
return $tool->execute($arguments, $companyId, $userId);
|
|
}
|
|
|
|
/**
|
|
* Test-only: reset the registry between tests that exercise different tool sets.
|
|
*/
|
|
public function flush(): void
|
|
{
|
|
$this->tools = [];
|
|
}
|
|
}
|