feat(training+controls): interactive video pipeline, training blocks, control generator, CE libraries

Interactive Training Videos (CP-TRAIN): - DB migration 022: training_checkpoints + checkpoint_progress tables - NarratorScript generation via Anthropic (AI Teacher persona, German) - TTS batch synthesis + interactive video pipeline (slides + checkpoint slides + FFmpeg) - 4 new API endpoints: generate-interactive, interactive-manifest, checkpoint submit, checkpoint progress - InteractiveVideoPlayer component (HTML5 Video, quiz overlay, seek protection, progress tracking) - Learner portal integration with automatic completion on all checkpoints passed - 30 new tests (handler validation + grading logic + manifest/progress + seek protection) Training Blocks: - Block generator, block store, block config CRUD + preview/generate endpoints - Migration 021: training_blocks schema Control Generator + Canonical Library: - Control generator routes + service enhancements - Canonical control library helpers, sidebar entry - Citation backfill service + tests - CE libraries data (hazard, protection, evidence, lifecycle, components) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-16 21:41:48 +01:00
parent d2133dbfa2
commit 4f6bc8f6f6
50 changed files with 17299 additions and 198 deletions
--- a/ai-compliance-sdk/internal/training/content_generator.go
+++ b/ai-compliance-sdk/internal/training/content_generator.go
@@ -294,6 +294,133 @@ func parseQuizResponse(response string, moduleID uuid.UUID) ([]QuizQuestion, err
 	return questions, nil
 }

+// GenerateBlockContent generates training content for a module based on linked canonical controls
+func (g *ContentGenerator) GenerateBlockContent(
+	ctx context.Context,
+	module TrainingModule,
+	controls []CanonicalControlSummary,
+	language string,
+) (*ModuleContent, error) {
+	if language == "" {
+		language = "de"
+	}
+
+	prompt := buildBlockContentPrompt(module, controls, language)
+
+	resp, err := g.registry.Chat(ctx, &llm.ChatRequest{
+		Messages: []llm.Message{
+			{Role: "system", Content: getContentSystemPrompt(language)},
+			{Role: "user", Content: prompt},
+		},
+		Temperature: 0.15,
+		MaxTokens:   8192,
+	})
+	if err != nil {
+		return nil, fmt.Errorf("LLM block content generation failed: %w", err)
+	}
+
+	contentBody := resp.Message.Content
+
+	// PII check
+	if g.piiDetector != nil && g.piiDetector.ContainsPII(contentBody) {
+		findings := g.piiDetector.FindPII(contentBody)
+		for _, f := range findings {
+			contentBody = strings.ReplaceAll(contentBody, f.Match, "[REDACTED]")
+		}
+	}
+
+	summary := contentBody
+	if len(summary) > 200 {
+		summary = summary[:200] + "..."
+	}
+
+	content := &ModuleContent{
+		ModuleID:      module.ID,
+		ContentFormat: ContentFormatMarkdown,
+		ContentBody:   contentBody,
+		Summary:       summary,
+		GeneratedBy:   "llm_block_" + resp.Provider,
+		LLMModel:      resp.Model,
+		IsPublished:   false,
+	}
+
+	if err := g.store.CreateModuleContent(ctx, content); err != nil {
+		return nil, fmt.Errorf("failed to save block content: %w", err)
+	}
+
+	// Audit log
+	g.store.LogAction(ctx, &AuditLogEntry{
+		TenantID:   module.TenantID,
+		Action:     AuditActionContentGenerated,
+		EntityType: AuditEntityModule,
+		EntityID:   &module.ID,
+		Details: map[string]interface{}{
+			"module_code":    module.ModuleCode,
+			"provider":       resp.Provider,
+			"model":          resp.Model,
+			"content_id":     content.ID.String(),
+			"version":        content.Version,
+			"tokens_used":    resp.Usage.TotalTokens,
+			"controls_count": len(controls),
+			"source":         "block_generator",
+		},
+	})
+
+	return content, nil
+}
+
+// buildBlockContentPrompt creates a prompt that incorporates canonical controls
+func buildBlockContentPrompt(module TrainingModule, controls []CanonicalControlSummary, language string) string {
+	var sb strings.Builder
+
+	if language == "en" {
+		sb.WriteString(fmt.Sprintf("Create training material for the following compliance module:\n\n"))
+		sb.WriteString(fmt.Sprintf("**Module Code:** %s\n", module.ModuleCode))
+		sb.WriteString(fmt.Sprintf("**Title:** %s\n", module.Title))
+		sb.WriteString(fmt.Sprintf("**Duration:** %d minutes\n\n", module.DurationMinutes))
+		sb.WriteString(fmt.Sprintf("This module is based on %d security controls:\n\n", len(controls)))
+	} else {
+		sb.WriteString(fmt.Sprintf("Erstelle Schulungsmaterial fuer folgendes Compliance-Modul:\n\n"))
+		sb.WriteString(fmt.Sprintf("**Modulcode:** %s\n", module.ModuleCode))
+		sb.WriteString(fmt.Sprintf("**Titel:** %s\n", module.Title))
+		sb.WriteString(fmt.Sprintf("**Dauer:** %d Minuten\n\n", module.DurationMinutes))
+		sb.WriteString(fmt.Sprintf("Dieses Modul basiert auf %d Sicherheits-Controls:\n\n", len(controls)))
+	}
+
+	for i, ctrl := range controls {
+		sb.WriteString(fmt.Sprintf("### Control %d: %s — %s\n", i+1, ctrl.ControlID, ctrl.Title))
+		sb.WriteString(fmt.Sprintf("**Ziel:** %s\n", ctrl.Objective))
+		if len(ctrl.Requirements) > 0 {
+			sb.WriteString("**Anforderungen:**\n")
+			for _, req := range ctrl.Requirements {
+				sb.WriteString(fmt.Sprintf("- %s\n", req))
+			}
+		}
+		sb.WriteString("\n")
+	}
+
+	if language == "en" {
+		sb.WriteString(`Create the material as Markdown:
+1. Introduction: Why are these controls important?
+2. Per control: Explanation, practical tips, examples
+3. Summary + action items
+4. Checklist for daily work
+
+Use clear, understandable language. Target audience: employees in companies (50-1,500 employees).`)
+	} else {
+		sb.WriteString(`Erstelle das Material als Markdown:
+1. Einfuehrung: Warum sind diese Controls wichtig?
+2. Pro Control: Erklaerung, praktische Hinweise, Beispiele
+3. Zusammenfassung + Handlungsanweisungen
+4. Checkliste fuer den Alltag
+
+Verwende klare, verstaendliche Sprache. Zielgruppe sind Mitarbeiter in Unternehmen (50-1.500 MA).
+Formatiere den Inhalt als Markdown mit Ueberschriften, Aufzaehlungen und Hervorhebungen.`)
+	}
+
+	return sb.String()
+}
+
 // GenerateAllModuleContent generates text content for all modules that don't have published content yet
 func (g *ContentGenerator) GenerateAllModuleContent(ctx context.Context, tenantID uuid.UUID, language string) (*BulkResult, error) {
 	if language == "" {
@@ -600,3 +727,252 @@ func truncateText(text string, maxLen int) string {
 	}
 	return text[:maxLen] + "..."
 }
+
+// ============================================================================
+// Interactive Video Pipeline
+// ============================================================================
+
+const narratorSystemPrompt = `Du bist ein professioneller AI Teacher fuer Compliance-Schulungen.
+Dein Stil ist foermlich aber freundlich, klar und paedagogisch wertvoll.
+Du sprichst die Lernenden direkt an ("Sie") und fuehrst sie durch die Schulung.
+Du erzeugst IMMER deutschsprachige Inhalte.
+
+Dein Output ist ein JSON-Objekt im Format NarratorScript.
+Jede Section sollte etwa 3 Minuten Sprechzeit haben (~450 Woerter Narrator-Text).
+Nach jeder Section kommt ein Checkpoint mit 3-5 Quiz-Fragen.
+Die Fragen testen das Verstaendnis des gerade Gelernten.
+Jede Frage hat genau 4 Antwortmoeglichkeiten, wobei correct_index (0-basiert) die richtige Antwort angibt.
+
+Antworte NUR mit dem JSON-Objekt, ohne Markdown-Codeblock-Wrapper.`
+
+// GenerateNarratorScript generates a narrator-style video script with checkpoints via LLM
+func (g *ContentGenerator) GenerateNarratorScript(ctx context.Context, module TrainingModule) (*NarratorScript, error) {
+	content, err := g.store.GetPublishedContent(ctx, module.ID)
+	if err != nil {
+		return nil, fmt.Errorf("failed to get content: %w", err)
+	}
+
+	contentContext := ""
+	if content != nil {
+		contentContext = fmt.Sprintf("\n\n**Vorhandener Schulungsinhalt (als Basis):**\n%s", truncateText(content.ContentBody, 4000))
+	}
+
+	prompt := fmt.Sprintf(`Erstelle ein interaktives Schulungsvideo-Skript mit Erzaehlerpersona und Checkpoints.
+
+**Modul:** %s — %s
+**Verordnung:** %s
+**Beschreibung:** %s
+**Dauer:** ca. %d Minuten
+%s
+
+Erstelle ein NarratorScript-JSON mit:
+- "title": Titel der Schulung
+- "intro": Begruessungstext ("Hallo, ich bin Ihr AI Teacher. Heute lernen Sie...")
+- "sections": Array mit 3-4 Abschnitten, jeder mit:
+  - "heading": Abschnittsueberschrift
+  - "narrator_text": Fliesstext im Erzaehlstil (~450 Woerter, ~3 Min Sprechzeit)
+  - "bullet_points": 3-5 Kernpunkte fuer die Folie
+  - "transition": Ueberleitung zum naechsten Abschnitt oder Checkpoint
+  - "checkpoint": Quiz-Block mit:
+    - "title": Checkpoint-Titel
+    - "questions": Array mit 3-5 Fragen, je:
+      - "question": Fragetext
+      - "options": Array mit 4 Antworten
+      - "correct_index": Index der richtigen Antwort (0-basiert)
+      - "explanation": Erklaerung der richtigen Antwort
+- "outro": Abschlussworte
+- "total_duration_estimate": geschaetzte Gesamtdauer in Sekunden
+
+Antworte NUR mit dem JSON-Objekt.`,
+		module.ModuleCode, module.Title,
+		string(module.RegulationArea),
+		module.Description,
+		module.DurationMinutes,
+		contentContext,
+	)
+
+	resp, err := g.registry.Chat(ctx, &llm.ChatRequest{
+		Messages: []llm.Message{
+			{Role: "system", Content: narratorSystemPrompt},
+			{Role: "user", Content: prompt},
+		},
+		Temperature: 0.2,
+		MaxTokens:   8192,
+	})
+	if err != nil {
+		return nil, fmt.Errorf("LLM narrator script generation failed: %w", err)
+	}
+
+	return parseNarratorScript(resp.Message.Content)
+}
+
+// parseNarratorScript extracts a NarratorScript from LLM output
+func parseNarratorScript(content string) (*NarratorScript, error) {
+	// Find JSON object in response
+	start := strings.Index(content, "{")
+	end := strings.LastIndex(content, "}")
+	if start < 0 || end <= start {
+		return nil, fmt.Errorf("no JSON object found in LLM response")
+	}
+	jsonStr := content[start : end+1]
+
+	var script NarratorScript
+	if err := json.Unmarshal([]byte(jsonStr), &script); err != nil {
+		return nil, fmt.Errorf("failed to parse narrator script JSON: %w", err)
+	}
+
+	if len(script.Sections) == 0 {
+		return nil, fmt.Errorf("narrator script has no sections")
+	}
+
+	return &script, nil
+}
+
+// GenerateInteractiveVideo orchestrates the full interactive video pipeline:
+// NarratorScript → TTS Audio → Slides+Video → DB Checkpoints + Quiz Questions
+func (g *ContentGenerator) GenerateInteractiveVideo(ctx context.Context, module TrainingModule) (*TrainingMedia, error) {
+	if g.ttsClient == nil {
+		return nil, fmt.Errorf("TTS client not configured")
+	}
+
+	// 1. Generate NarratorScript via LLM
+	script, err := g.GenerateNarratorScript(ctx, module)
+	if err != nil {
+		return nil, fmt.Errorf("narrator script generation failed: %w", err)
+	}
+
+	// 2. Synthesize audio per section via TTS service
+	sections := make([]SectionAudio, len(script.Sections))
+	for i, s := range script.Sections {
+		// Combine narrator text with intro/outro for first/last section
+		text := s.NarratorText
+		if i == 0 && script.Intro != "" {
+			text = script.Intro + "\n\n" + text
+		}
+		if i == len(script.Sections)-1 && script.Outro != "" {
+			text = text + "\n\n" + script.Outro
+		}
+		sections[i] = SectionAudio{
+			Text:    text,
+			Heading: s.Heading,
+		}
+	}
+
+	audioResp, err := g.ttsClient.SynthesizeSections(ctx, &SynthesizeSectionsRequest{
+		Sections: sections,
+		Voice:    "de_DE-thorsten-high",
+		ModuleID: module.ID.String(),
+	})
+	if err != nil {
+		return nil, fmt.Errorf("section audio synthesis failed: %w", err)
+	}
+
+	// 3. Generate interactive video via TTS service
+	videoResp, err := g.ttsClient.GenerateInteractiveVideo(ctx, &GenerateInteractiveVideoRequest{
+		Script:   script,
+		Audio:    audioResp,
+		ModuleID: module.ID.String(),
+	})
+	if err != nil {
+		return nil, fmt.Errorf("interactive video generation failed: %w", err)
+	}
+
+	// 4. Save TrainingMedia record
+	scriptJSON, _ := json.Marshal(script)
+	media := &TrainingMedia{
+		ModuleID:    module.ID,
+		MediaType:   MediaTypeInteractiveVideo,
+		Status:      MediaStatusProcessing,
+		Bucket:      "compliance-training-video",
+		ObjectKey:   fmt.Sprintf("video/%s/interactive.mp4", module.ID.String()),
+		MimeType:    "video/mp4",
+		Language:    "de",
+		GeneratedBy: "tts_ffmpeg_interactive",
+		Metadata:    scriptJSON,
+	}
+
+	if err := g.store.CreateMedia(ctx, media); err != nil {
+		return nil, fmt.Errorf("failed to create media record: %w", err)
+	}
+
+	// Update media with video result
+	media.Status = MediaStatusCompleted
+	media.FileSizeBytes = videoResp.SizeBytes
+	media.DurationSeconds = videoResp.DurationSeconds
+	media.ObjectKey = videoResp.ObjectKey
+	media.Bucket = videoResp.Bucket
+	g.store.UpdateMediaStatus(ctx, media.ID, MediaStatusCompleted, videoResp.SizeBytes, videoResp.DurationSeconds, "")
+
+	// Auto-publish
+	g.store.PublishMedia(ctx, media.ID, true)
+
+	// 5. Create Checkpoints + Quiz Questions in DB
+	// Clear old checkpoints first
+	g.store.DeleteCheckpointsForModule(ctx, module.ID)
+
+	for i, section := range script.Sections {
+		if section.Checkpoint == nil {
+			continue
+		}
+
+		// Calculate timestamp from cumulative audio durations
+		var timestamp float64
+		if i < len(audioResp.Sections) {
+			// Checkpoint timestamp = end of this section's audio
+			timestamp = audioResp.Sections[i].StartTimestamp + audioResp.Sections[i].Duration
+		}
+
+		cp := &Checkpoint{
+			ModuleID:         module.ID,
+			CheckpointIndex:  i,
+			Title:            section.Checkpoint.Title,
+			TimestampSeconds: timestamp,
+		}
+		if err := g.store.CreateCheckpoint(ctx, cp); err != nil {
+			return nil, fmt.Errorf("failed to create checkpoint %d: %w", i, err)
+		}
+
+		// Save quiz questions for this checkpoint
+		for j, q := range section.Checkpoint.Questions {
+			question := &QuizQuestion{
+				ModuleID:     module.ID,
+				Question:     q.Question,
+				Options:      q.Options,
+				CorrectIndex: q.CorrectIndex,
+				Explanation:  q.Explanation,
+				Difficulty:   DifficultyMedium,
+				SortOrder:    j,
+			}
+			if err := g.store.CreateCheckpointQuizQuestion(ctx, question, cp.ID); err != nil {
+				return nil, fmt.Errorf("failed to create checkpoint question: %w", err)
+			}
+		}
+	}
+
+	// 6. Audit log
+	g.store.LogAction(ctx, &AuditLogEntry{
+		TenantID:   module.TenantID,
+		Action:     AuditAction("interactive_video_generated"),
+		EntityType: AuditEntityModule,
+		EntityID:   &module.ID,
+		Details: map[string]interface{}{
+			"module_code":      module.ModuleCode,
+			"media_id":         media.ID.String(),
+			"duration_seconds": videoResp.DurationSeconds,
+			"sections":         len(script.Sections),
+			"checkpoints":      countCheckpoints(script),
+		},
+	})
+
+	return media, nil
+}
+
+func countCheckpoints(script *NarratorScript) int {
+	count := 0
+	for _, s := range script.Sections {
+		if s.Checkpoint != nil {
+			count++
+		}
+	}
+	return count
+}