Add configurable export error handling policies (bd-exug)

Implements flexible error handling for export operations with four policies:
- strict: Fail-fast on any error (default for user exports)
- best-effort: Skip errors with warnings (default for auto-exports)
- partial: Retry then skip with manifest tracking
- required-core: Fail on core data, skip enrichments

Key features:
- Per-project configuration via `bd config set export.error_policy`
- Separate policy for auto-exports: `auto_export.error_policy`
- Retry with exponential backoff (configurable attempts/delay)
- Optional export manifests documenting completeness
- Per-issue encoding error handling

This allows users to choose the right trade-off between data integrity
and system availability for their specific project needs.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

internal/export/policy_test.go

@@ -0,0 +1,176 @@
+package export
+
+import (
+	"context"
+	"errors"
+	"testing"
+	"time"
+)
+
+func TestRetryWithBackoff(t *testing.T) {
+	ctx := context.Background()
+
+	t.Run("succeeds first try", func(t *testing.T) {
+		attempts := 0
+		err := RetryWithBackoff(ctx, 3, 100, "test", func() error {
+			attempts++
+			return nil
+		})
+		if err != nil {
+			t.Errorf("expected no error, got %v", err)
+		}
+		if attempts != 1 {
+			t.Errorf("expected 1 attempt, got %d", attempts)
+		}
+	})
+
+	t.Run("succeeds after retries", func(t *testing.T) {
+		attempts := 0
+		err := RetryWithBackoff(ctx, 3, 10, "test", func() error {
+			attempts++
+			if attempts < 3 {
+				return errors.New("transient error")
+			}
+			return nil
+		})
+		if err != nil {
+			t.Errorf("expected no error, got %v", err)
+		}
+		if attempts != 3 {
+			t.Errorf("expected 3 attempts, got %d", attempts)
+		}
+	})
+
+	t.Run("fails after max retries", func(t *testing.T) {
+		attempts := 0
+		err := RetryWithBackoff(ctx, 3, 10, "test", func() error {
+			attempts++
+			return errors.New("persistent error")
+		})
+		if err == nil {
+			t.Error("expected error, got nil")
+		}
+		if attempts != 3 {
+			t.Errorf("expected 3 attempts, got %d", attempts)
+		}
+	})
+
+	t.Run("respects context cancellation", func(t *testing.T) {
+		ctx, cancel := context.WithTimeout(context.Background(), 50*time.Millisecond)
+		defer cancel()
+
+		attempts := 0
+		err := RetryWithBackoff(ctx, 10, 100, "test", func() error {
+			attempts++
+			return errors.New("error")
+		})
+		if err != context.DeadlineExceeded {
+			t.Errorf("expected DeadlineExceeded, got %v", err)
+		}
+		// Should stop before reaching max retries due to timeout
+		if attempts >= 10 {
+			t.Errorf("expected fewer than 10 attempts due to timeout, got %d", attempts)
+		}
+	})
+}
+
+func TestErrorPolicy(t *testing.T) {
+	tests := []struct {
+		name  string
+		policy ErrorPolicy
+		valid bool
+	}{
+		{"strict", PolicyStrict, true},
+		{"best-effort", PolicyBestEffort, true},
+		{"partial", PolicyPartial, true},
+		{"required-core", PolicyRequiredCore, true},
+		{"invalid", ErrorPolicy("invalid"), false},
+		{"empty", ErrorPolicy(""), false},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := tt.policy.IsValid(); got != tt.valid {
+				t.Errorf("IsValid() = %v, want %v", got, tt.valid)
+			}
+		})
+	}
+}
+
+func TestFetchWithPolicy(t *testing.T) {
+	ctx := context.Background()
+
+	t.Run("strict policy fails fast", func(t *testing.T) {
+		cfg := &Config{
+			Policy:         PolicyStrict,
+			RetryAttempts:  1,
+			RetryBackoffMS: 10,
+		}
+		result := FetchWithPolicy(ctx, cfg, DataTypeCore, "test", func() error {
+			return errors.New("test error")
+		})
+		if result.Err == nil {
+			t.Error("expected error, got nil")
+		}
+		if result.Success {
+			t.Error("expected Success=false")
+		}
+	})
+
+	t.Run("best-effort policy skips errors", func(t *testing.T) {
+		cfg := &Config{
+			Policy:         PolicyBestEffort,
+			RetryAttempts:  1,
+			RetryBackoffMS: 10,
+		}
+		result := FetchWithPolicy(ctx, cfg, DataTypeLabels, "test", func() error {
+			return errors.New("test error")
+		})
+		if result.Err != nil {
+			t.Errorf("expected no error in best-effort, got %v", result.Err)
+		}
+		if result.Success {
+			t.Error("expected Success=false")
+		}
+		if len(result.Warnings) == 0 {
+			t.Error("expected warnings")
+		}
+	})
+
+	t.Run("required-core fails on core data", func(t *testing.T) {
+		cfg := &Config{
+			Policy:         PolicyRequiredCore,
+			RetryAttempts:  1,
+			RetryBackoffMS: 10,
+		}
+		result := FetchWithPolicy(ctx, cfg, DataTypeCore, "test", func() error {
+			return errors.New("test error")
+		})
+		if result.Err == nil {
+			t.Error("expected error for core data, got nil")
+		}
+		if result.Success {
+			t.Error("expected Success=false")
+		}
+	})
+
+	t.Run("required-core skips enrichment errors", func(t *testing.T) {
+		cfg := &Config{
+			Policy:         PolicyRequiredCore,
+			RetryAttempts:  1,
+			RetryBackoffMS: 10,
+		}
+		result := FetchWithPolicy(ctx, cfg, DataTypeLabels, "test", func() error {
+			return errors.New("test error")
+		})
+		if result.Err != nil {
+			t.Errorf("expected no error for enrichment, got %v", result.Err)
+		}
+		if result.Success {
+			t.Error("expected Success=false")
+		}
+		if len(result.Warnings) == 0 {
+			t.Error("expected warnings")
+		}
+	})
+}