feat(11-01): add DuckDuckGoSource, YandexSource, and BraveSource

- DuckDuckGoSource scrapes HTML search (no API key, always enabled, RespectsRobots=true)
- YandexSource uses Yandex XML Search API (user+key required, XML response parsing)
- BraveSource uses Brave Search API (X-Subscription-Token header, JSON response)
- All three follow established error handling: 401 aborts, transient continues, ctx cancellation returns

pkg/recon/sources/brave.go

@@ -0,0 +1,153 @@
+package sources
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+	"net/url"
+	"strings"
+	"time"
+
+	"golang.org/x/time/rate"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+// BraveSource implements recon.ReconSource against the Brave Search API.
+// It requires an API key (X-Subscription-Token) to be enabled.
+type BraveSource struct {
+	APIKey   string
+	BaseURL  string
+	Registry *providers.Registry
+	Limiters *recon.LimiterRegistry
+	client   *Client
+}
+
+// Compile-time assertion.
+var _ recon.ReconSource = (*BraveSource)(nil)
+
+// NewBraveSource constructs a BraveSource with the shared retry client.
+func NewBraveSource(apiKey string, reg *providers.Registry, lim *recon.LimiterRegistry) *BraveSource {
+	return &BraveSource{
+		APIKey:   apiKey,
+		BaseURL:  "https://api.search.brave.com",
+		Registry: reg,
+		Limiters: lim,
+		client:   NewClient(),
+	}
+}
+
+func (s *BraveSource) Name() string          { return "brave" }
+func (s *BraveSource) RateLimit() rate.Limit { return rate.Every(1 * time.Second) }
+func (s *BraveSource) Burst() int            { return 1 }
+func (s *BraveSource) RespectsRobots() bool  { return false }
+
+// Enabled returns true only when APIKey is configured.
+func (s *BraveSource) Enabled(_ recon.Config) bool { return s.APIKey != "" }
+
+// Sweep issues one Brave Search request per provider keyword and emits a
+// Finding for every web result.
+func (s *BraveSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
+	if s.APIKey == "" {
+		return nil
+	}
+	base := s.BaseURL
+	if base == "" {
+		base = "https://api.search.brave.com"
+	}
+
+	queries := BuildQueries(s.Registry, "brave")
+	kwIndex := braveKeywordIndex(s.Registry)
+
+	for _, q := range queries {
+		if err := ctx.Err(); err != nil {
+			return err
+		}
+		if s.Limiters != nil {
+			if err := s.Limiters.Wait(ctx, s.Name(), s.RateLimit(), s.Burst(), false); err != nil {
+				return err
+			}
+		}
+
+		endpoint := fmt.Sprintf("%s/res/v1/web/search?q=%s&count=20", base, url.QueryEscape(q))
+		req, err := http.NewRequestWithContext(ctx, http.MethodGet, endpoint, nil)
+		if err != nil {
+			return fmt.Errorf("brave: build request: %w", err)
+		}
+		req.Header.Set("X-Subscription-Token", s.APIKey)
+		req.Header.Set("Accept", "application/json")
+		req.Header.Set("User-Agent", "keyhunter-recon")
+
+		resp, err := s.client.Do(ctx, req)
+		if err != nil {
+			if errors.Is(err, ErrUnauthorized) {
+				return err
+			}
+			if errors.Is(err, context.Canceled) || errors.Is(err, context.DeadlineExceeded) {
+				return err
+			}
+			continue
+		}
+
+		var parsed braveSearchResponse
+		decErr := json.NewDecoder(resp.Body).Decode(&parsed)
+		_ = resp.Body.Close()
+		if decErr != nil {
+			continue
+		}
+
+		provName := kwIndex[strings.ToLower(extractGoogleKeyword(q))]
+		for _, it := range parsed.Web.Results {
+			f := recon.Finding{
+				ProviderName: provName,
+				Confidence:   "low",
+				Source:       it.URL,
+				SourceType:   "recon:brave",
+				DetectedAt:   time.Now(),
+			}
+			select {
+			case out <- f:
+			case <-ctx.Done():
+				return ctx.Err()
+			}
+		}
+	}
+	return nil
+}
+
+type braveSearchResponse struct {
+	Web braveWebResults `json:"web"`
+}
+
+type braveWebResults struct {
+	Results []braveWebItem `json:"results"`
+}
+
+type braveWebItem struct {
+	URL         string `json:"url"`
+	Title       string `json:"title"`
+	Description string `json:"description"`
+}
+
+// braveKeywordIndex maps lowercased keywords to provider names.
+func braveKeywordIndex(reg *providers.Registry) map[string]string {
+	m := make(map[string]string)
+	if reg == nil {
+		return m
+	}
+	for _, p := range reg.List() {
+		for _, k := range p.Keywords {
+			kl := strings.ToLower(strings.TrimSpace(k))
+			if kl == "" {
+				continue
+			}
+			if _, exists := m[kl]; !exists {
+				m[kl] = p.Name
+			}
+		}
+	}
+	return m
+}

pkg/recon/sources/brave_test.go

@@ -0,0 +1,145 @@
+package sources
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+func braveStubHandler(t *testing.T, calls *int32) http.HandlerFunc {
+	t.Helper()
+	return func(w http.ResponseWriter, r *http.Request) {
+		atomic.AddInt32(calls, 1)
+		if !strings.HasPrefix(r.URL.Path, "/res/v1/web/search") {
+			t.Errorf("unexpected path: %s", r.URL.Path)
+		}
+		if got := r.Header.Get("X-Subscription-Token"); got != "testtoken" {
+			t.Errorf("missing subscription token: %q", got)
+		}
+		body := map[string]any{
+			"web": map[string]any{
+				"results": []map[string]any{
+					{"url": "https://pastebin.com/brave1", "title": "Brave Result 1", "description": "found key"},
+					{"url": "https://github.com/org/repo/blob/main/config.env", "title": "Brave Result 2", "description": "leaked"},
+				},
+			},
+		}
+		w.Header().Set("Content-Type", "application/json")
+		_ = json.NewEncoder(w).Encode(body)
+	}
+}
+
+func TestBraveSource_EnabledRequiresAPIKey(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+
+	if s := NewBraveSource("", reg, lim); s.Enabled(recon.Config{}) {
+		t.Error("expected Enabled=false with empty key")
+	}
+	if s := NewBraveSource("key", reg, lim); !s.Enabled(recon.Config{}) {
+		t.Error("expected Enabled=true with key")
+	}
+}
+
+func TestBraveSource_SweepEmptyKeyReturnsNil(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	s := NewBraveSource("", reg, lim)
+
+	out := make(chan recon.Finding, 10)
+	if err := s.Sweep(context.Background(), "", out); err != nil {
+		t.Fatalf("expected nil, got %v", err)
+	}
+	close(out)
+	if n := countFindings(out); n != 0 {
+		t.Fatalf("expected 0 findings, got %d", n)
+	}
+}
+
+func TestBraveSource_SweepEmitsFindings(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("brave", 1000, 100)
+
+	var calls int32
+	srv := httptest.NewServer(braveStubHandler(t, &calls))
+	defer srv.Close()
+
+	s := NewBraveSource("testtoken", reg, lim)
+	s.BaseURL = srv.URL
+
+	out := make(chan recon.Finding, 32)
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+
+	done := make(chan error, 1)
+	go func() { done <- s.Sweep(ctx, "", out); close(out) }()
+
+	var findings []recon.Finding
+	for f := range out {
+		findings = append(findings, f)
+	}
+	if err := <-done; err != nil {
+		t.Fatalf("Sweep error: %v", err)
+	}
+
+	// 2 keywords * 2 items = 4 findings
+	if len(findings) != 4 {
+		t.Fatalf("expected 4 findings, got %d", len(findings))
+	}
+	for _, f := range findings {
+		if f.SourceType != "recon:brave" {
+			t.Errorf("SourceType=%q want recon:brave", f.SourceType)
+		}
+	}
+	if got := atomic.LoadInt32(&calls); got != 2 {
+		t.Errorf("expected 2 calls, got %d", got)
+	}
+}
+
+func TestBraveSource_CtxCancelled(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("brave", 1000, 100)
+
+	s := NewBraveSource("key", reg, lim)
+	s.BaseURL = "http://127.0.0.1:1"
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	out := make(chan recon.Finding, 1)
+	err := s.Sweep(ctx, "", out)
+	if !errors.Is(err, context.Canceled) {
+		t.Fatalf("expected context.Canceled, got %v", err)
+	}
+}
+
+func TestBraveSource_Unauthorized(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("brave", 1000, 100)
+
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.WriteHeader(http.StatusUnauthorized)
+		_, _ = w.Write([]byte("bad token"))
+	}))
+	defer srv.Close()
+
+	s := NewBraveSource("key", reg, lim)
+	s.BaseURL = srv.URL
+
+	out := make(chan recon.Finding, 1)
+	err := s.Sweep(context.Background(), "", out)
+	if !errors.Is(err, ErrUnauthorized) {
+		t.Fatalf("expected ErrUnauthorized, got %v", err)
+	}
+}

pkg/recon/sources/duckduckgo.go

@@ -0,0 +1,116 @@
+package sources
+
+import (
+	"context"
+	"fmt"
+	"net/http"
+	"net/url"
+	"regexp"
+	"time"
+
+	"golang.org/x/time/rate"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+// DuckDuckGoSource implements recon.ReconSource by scraping DuckDuckGo's HTML
+// search endpoint. No API key is required -- this source is always enabled.
+//
+// It operates conservatively (2s per request) and declares RespectsRobots=true.
+type DuckDuckGoSource struct {
+	BaseURL  string
+	Registry *providers.Registry
+	Limiters *recon.LimiterRegistry
+	client   *Client
+}
+
+// Compile-time assertion.
+var _ recon.ReconSource = (*DuckDuckGoSource)(nil)
+
+// ddgResultRE matches DuckDuckGo HTML result links. The HTML search page uses
+// <a class="result__a" href="..."> anchors for organic results.
+var ddgResultRE = regexp.MustCompile(`^https?://`)
+
+// NewDuckDuckGoSource constructs a DuckDuckGoSource with the shared retry client.
+func NewDuckDuckGoSource(reg *providers.Registry, lim *recon.LimiterRegistry) *DuckDuckGoSource {
+	return &DuckDuckGoSource{
+		BaseURL:  "https://html.duckduckgo.com",
+		Registry: reg,
+		Limiters: lim,
+		client:   NewClient(),
+	}
+}
+
+func (s *DuckDuckGoSource) Name() string          { return "duckduckgo" }
+func (s *DuckDuckGoSource) RateLimit() rate.Limit { return rate.Every(2 * time.Second) }
+func (s *DuckDuckGoSource) Burst() int            { return 1 }
+func (s *DuckDuckGoSource) RespectsRobots() bool  { return true }
+
+// Enabled always returns true -- DuckDuckGo HTML scraping requires no credentials.
+func (s *DuckDuckGoSource) Enabled(_ recon.Config) bool { return true }
+
+// Sweep iterates provider keywords, scrapes DuckDuckGo HTML search, and emits
+// a Finding per result link.
+func (s *DuckDuckGoSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
+	base := s.BaseURL
+	if base == "" {
+		base = "https://html.duckduckgo.com"
+	}
+	client := s.client
+	if client == nil {
+		client = NewClient()
+	}
+
+	queries := BuildQueries(s.Registry, "duckduckgo")
+	if len(queries) == 0 {
+		return nil
+	}
+
+	for _, q := range queries {
+		if err := ctx.Err(); err != nil {
+			return err
+		}
+		if s.Limiters != nil {
+			if err := s.Limiters.Wait(ctx, s.Name(), s.RateLimit(), s.Burst(), false); err != nil {
+				return err
+			}
+		}
+
+		searchURL := fmt.Sprintf("%s/html/?q=%s", base, url.QueryEscape(q))
+		req, err := http.NewRequestWithContext(ctx, http.MethodGet, searchURL, nil)
+		if err != nil {
+			return fmt.Errorf("duckduckgo: build req: %w", err)
+		}
+		req.Header.Set("User-Agent", "keyhunter-recon")
+
+		resp, err := client.Do(ctx, req)
+		if err != nil {
+			// Transient failures: continue to next query.
+			continue
+		}
+		links, parseErr := extractAnchorHrefs(resp.Body, ddgResultRE)
+		_ = resp.Body.Close()
+		if parseErr != nil {
+			continue
+		}
+
+		for _, href := range links {
+			if err := ctx.Err(); err != nil {
+				return err
+			}
+			f := recon.Finding{
+				Source:     href,
+				SourceType: "recon:duckduckgo",
+				Confidence: "low",
+				DetectedAt: time.Now(),
+			}
+			select {
+			case out <- f:
+			case <-ctx.Done():
+				return ctx.Err()
+			}
+		}
+	}
+	return nil
+}

pkg/recon/sources/duckduckgo_test.go

@@ -0,0 +1,134 @@
+package sources
+
+import (
+	"context"
+	"errors"
+	"net/http"
+	"net/http/httptest"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+const ddgHTMLFixture = `<!DOCTYPE html>
+<html>
+<body>
+<div class="results">
+  <div class="result">
+    <a class="result__a" href="https://pastebin.com/abc123">Pastebin Result</a>
+  </div>
+  <div class="result">
+    <a class="result__a" href="https://github.com/user/repo/blob/main/.env">GitHub Result</a>
+  </div>
+  <div class="result">
+    <a class="result__a" href="https://example.com/page">Example</a>
+  </div>
+</div>
+</body>
+</html>`
+
+func ddgStubHandler(t *testing.T, calls *int32) http.HandlerFunc {
+	t.Helper()
+	return func(w http.ResponseWriter, r *http.Request) {
+		atomic.AddInt32(calls, 1)
+		if r.URL.Path != "/html/" {
+			t.Errorf("unexpected path: %s", r.URL.Path)
+		}
+		w.Header().Set("Content-Type", "text/html")
+		_, _ = w.Write([]byte(ddgHTMLFixture))
+	}
+}
+
+func TestDuckDuckGoSource_AlwaysEnabled(t *testing.T) {
+	s := NewDuckDuckGoSource(syntheticRegistry(), recon.NewLimiterRegistry())
+	if !s.Enabled(recon.Config{}) {
+		t.Error("expected Enabled=true always")
+	}
+}
+
+func TestDuckDuckGoSource_RespectsRobots(t *testing.T) {
+	s := NewDuckDuckGoSource(syntheticRegistry(), recon.NewLimiterRegistry())
+	if !s.RespectsRobots() {
+		t.Error("expected RespectsRobots=true")
+	}
+}
+
+func TestDuckDuckGoSource_SweepEmitsFindings(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("duckduckgo", 1000, 100)
+
+	var calls int32
+	srv := httptest.NewServer(ddgStubHandler(t, &calls))
+	defer srv.Close()
+
+	s := NewDuckDuckGoSource(reg, lim)
+	s.BaseURL = srv.URL
+
+	out := make(chan recon.Finding, 32)
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+
+	done := make(chan error, 1)
+	go func() { done <- s.Sweep(ctx, "", out); close(out) }()
+
+	var findings []recon.Finding
+	for f := range out {
+		findings = append(findings, f)
+	}
+	if err := <-done; err != nil {
+		t.Fatalf("Sweep error: %v", err)
+	}
+
+	// 2 keywords * 3 links = 6 findings
+	if len(findings) != 6 {
+		t.Fatalf("expected 6 findings, got %d", len(findings))
+	}
+	for _, f := range findings {
+		if f.SourceType != "recon:duckduckgo" {
+			t.Errorf("SourceType=%q want recon:duckduckgo", f.SourceType)
+		}
+		if f.Confidence != "low" {
+			t.Errorf("Confidence=%q want low", f.Confidence)
+		}
+	}
+	if got := atomic.LoadInt32(&calls); got != 2 {
+		t.Errorf("expected 2 DDG calls, got %d", got)
+	}
+}
+
+func TestDuckDuckGoSource_CtxCancelled(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("duckduckgo", 1000, 100)
+
+	s := NewDuckDuckGoSource(reg, lim)
+	s.BaseURL = "http://127.0.0.1:1"
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	out := make(chan recon.Finding, 1)
+	err := s.Sweep(ctx, "", out)
+	if !errors.Is(err, context.Canceled) {
+		t.Fatalf("expected context.Canceled, got %v", err)
+	}
+}
+
+func TestDuckDuckGoSource_EmptyRegistryNoError(t *testing.T) {
+	lim := recon.NewLimiterRegistry()
+	s := NewDuckDuckGoSource(nil, lim)
+
+	out := make(chan recon.Finding, 1)
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		t.Fatal("should not be called with nil registry")
+	}))
+	defer srv.Close()
+	s.BaseURL = srv.URL
+
+	if err := s.Sweep(context.Background(), "", out); err != nil {
+		t.Fatalf("expected nil, got %v", err)
+	}
+}

pkg/recon/sources/yandex.go

@@ -0,0 +1,177 @@
+package sources
+
+import (
+	"context"
+	"encoding/xml"
+	"errors"
+	"fmt"
+	"net/http"
+	"net/url"
+	"strings"
+	"time"
+
+	"golang.org/x/time/rate"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+// YandexSource implements recon.ReconSource against the Yandex XML Search API.
+// It requires both a User and APIKey to be enabled.
+type YandexSource struct {
+	User     string
+	APIKey   string
+	BaseURL  string
+	Registry *providers.Registry
+	Limiters *recon.LimiterRegistry
+	client   *Client
+}
+
+// Compile-time assertion.
+var _ recon.ReconSource = (*YandexSource)(nil)
+
+// NewYandexSource constructs a YandexSource with the shared retry client.
+func NewYandexSource(user, apiKey string, reg *providers.Registry, lim *recon.LimiterRegistry) *YandexSource {
+	return &YandexSource{
+		User:     user,
+		APIKey:   apiKey,
+		BaseURL:  "https://yandex.com",
+		Registry: reg,
+		Limiters: lim,
+		client:   NewClient(),
+	}
+}
+
+func (s *YandexSource) Name() string          { return "yandex" }
+func (s *YandexSource) RateLimit() rate.Limit { return rate.Every(1 * time.Second) }
+func (s *YandexSource) Burst() int            { return 1 }
+func (s *YandexSource) RespectsRobots() bool  { return false }
+
+// Enabled returns true only when both User and APIKey are configured.
+func (s *YandexSource) Enabled(_ recon.Config) bool {
+	return s.User != "" && s.APIKey != ""
+}
+
+// Sweep issues one Yandex XML search request per provider keyword and emits a
+// Finding for every <url> element in the response.
+func (s *YandexSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
+	if s.User == "" || s.APIKey == "" {
+		return nil
+	}
+	base := s.BaseURL
+	if base == "" {
+		base = "https://yandex.com"
+	}
+
+	queries := BuildQueries(s.Registry, "yandex")
+	kwIndex := yandexKeywordIndex(s.Registry)
+
+	for _, q := range queries {
+		if err := ctx.Err(); err != nil {
+			return err
+		}
+		if s.Limiters != nil {
+			if err := s.Limiters.Wait(ctx, s.Name(), s.RateLimit(), s.Burst(), false); err != nil {
+				return err
+			}
+		}
+
+		endpoint := fmt.Sprintf("%s/search/xml?user=%s&key=%s&query=%s&l10n=en&sortby=rlv&filter=none&groupby=%s",
+			base,
+			url.QueryEscape(s.User),
+			url.QueryEscape(s.APIKey),
+			url.QueryEscape(q),
+			url.QueryEscape(`attr="".mode=flat.groups-on-page=50`))
+		req, err := http.NewRequestWithContext(ctx, http.MethodGet, endpoint, nil)
+		if err != nil {
+			return fmt.Errorf("yandex: build request: %w", err)
+		}
+		req.Header.Set("User-Agent", "keyhunter-recon")
+
+		resp, err := s.client.Do(ctx, req)
+		if err != nil {
+			if errors.Is(err, ErrUnauthorized) {
+				return err
+			}
+			if errors.Is(err, context.Canceled) || errors.Is(err, context.DeadlineExceeded) {
+				return err
+			}
+			continue
+		}
+
+		var parsed yandexSearchResponse
+		decErr := xml.NewDecoder(resp.Body).Decode(&parsed)
+		_ = resp.Body.Close()
+		if decErr != nil {
+			continue
+		}
+
+		provName := kwIndex[strings.ToLower(extractGoogleKeyword(q))]
+		for _, grp := range parsed.Response.Results.Grouping.Groups {
+			for _, doc := range grp.Docs {
+				if doc.URL == "" {
+					continue
+				}
+				f := recon.Finding{
+					ProviderName: provName,
+					Confidence:   "low",
+					Source:       doc.URL,
+					SourceType:   "recon:yandex",
+					DetectedAt:   time.Now(),
+				}
+				select {
+				case out <- f:
+				case <-ctx.Done():
+					return ctx.Err()
+				}
+			}
+		}
+	}
+	return nil
+}
+
+// XML response structures for Yandex XML Search API.
+type yandexSearchResponse struct {
+	XMLName  xml.Name       `xml:"yandexsearch"`
+	Response yandexResponse `xml:"response"`
+}
+
+type yandexResponse struct {
+	Results yandexResults `xml:"results"`
+}
+
+type yandexResults struct {
+	Grouping yandexGrouping `xml:"grouping"`
+}
+
+type yandexGrouping struct {
+	Groups []yandexGroup `xml:"group"`
+}
+
+type yandexGroup struct {
+	Docs []yandexDoc `xml:"doc"`
+}
+
+type yandexDoc struct {
+	URL string `xml:"url"`
+}
+
+// yandexKeywordIndex maps lowercased keywords to provider names.
+func yandexKeywordIndex(reg *providers.Registry) map[string]string {
+	m := make(map[string]string)
+	if reg == nil {
+		return m
+	}
+	for _, p := range reg.List() {
+		for _, k := range p.Keywords {
+			kl := strings.ToLower(strings.TrimSpace(k))
+			if kl == "" {
+				continue
+			}
+			if _, exists := m[kl]; !exists {
+				m[kl] = p.Name
+			}
+		}
+	}
+	return m
+}

pkg/recon/sources/yandex_test.go

@@ -0,0 +1,171 @@
+package sources
+
+import (
+	"context"
+	"errors"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+const yandexXMLFixture = `<?xml version="1.0" encoding="utf-8"?>
+<yandexsearch>
+  <response>
+    <results>
+      <grouping>
+        <group>
+          <doc>
+            <url>https://pastebin.com/yandex1</url>
+          </doc>
+        </group>
+        <group>
+          <doc>
+            <url>https://github.com/user/repo/blob/main/secrets.env</url>
+          </doc>
+          <doc>
+            <url>https://example.com/leaked</url>
+          </doc>
+        </group>
+      </grouping>
+    </results>
+  </response>
+</yandexsearch>`
+
+func yandexStubHandler(t *testing.T, calls *int32) http.HandlerFunc {
+	t.Helper()
+	return func(w http.ResponseWriter, r *http.Request) {
+		atomic.AddInt32(calls, 1)
+		if !strings.HasPrefix(r.URL.Path, "/search/xml") {
+			t.Errorf("unexpected path: %s", r.URL.Path)
+		}
+		if r.URL.Query().Get("user") != "testuser" {
+			t.Errorf("missing user param")
+		}
+		if r.URL.Query().Get("key") != "testkey" {
+			t.Errorf("missing key param")
+		}
+		w.Header().Set("Content-Type", "application/xml")
+		_, _ = w.Write([]byte(yandexXMLFixture))
+	}
+}
+
+func TestYandexSource_EnabledRequiresBoth(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+
+	tests := []struct {
+		user, key string
+		want      bool
+	}{
+		{"", "", false},
+		{"user", "", false},
+		{"", "key", false},
+		{"user", "key", true},
+	}
+	for _, tt := range tests {
+		s := NewYandexSource(tt.user, tt.key, reg, lim)
+		if got := s.Enabled(recon.Config{}); got != tt.want {
+			t.Errorf("Enabled(user=%q, key=%q) = %v, want %v", tt.user, tt.key, got, tt.want)
+		}
+	}
+}
+
+func TestYandexSource_SweepEmptyCredsReturnsNil(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	s := NewYandexSource("", "", reg, lim)
+
+	out := make(chan recon.Finding, 10)
+	if err := s.Sweep(context.Background(), "", out); err != nil {
+		t.Fatalf("expected nil, got %v", err)
+	}
+	close(out)
+	if n := countFindings(out); n != 0 {
+		t.Fatalf("expected 0 findings, got %d", n)
+	}
+}
+
+func TestYandexSource_SweepEmitsFindings(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("yandex", 1000, 100)
+
+	var calls int32
+	srv := httptest.NewServer(yandexStubHandler(t, &calls))
+	defer srv.Close()
+
+	s := NewYandexSource("testuser", "testkey", reg, lim)
+	s.BaseURL = srv.URL
+
+	out := make(chan recon.Finding, 32)
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+
+	done := make(chan error, 1)
+	go func() { done <- s.Sweep(ctx, "", out); close(out) }()
+
+	var findings []recon.Finding
+	for f := range out {
+		findings = append(findings, f)
+	}
+	if err := <-done; err != nil {
+		t.Fatalf("Sweep error: %v", err)
+	}
+
+	// 2 keywords * 3 URLs in XML = 6 findings
+	if len(findings) != 6 {
+		t.Fatalf("expected 6 findings, got %d", len(findings))
+	}
+	for _, f := range findings {
+		if f.SourceType != "recon:yandex" {
+			t.Errorf("SourceType=%q want recon:yandex", f.SourceType)
+		}
+	}
+	if got := atomic.LoadInt32(&calls); got != 2 {
+		t.Errorf("expected 2 calls, got %d", got)
+	}
+}
+
+func TestYandexSource_CtxCancelled(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("yandex", 1000, 100)
+
+	s := NewYandexSource("user", "key", reg, lim)
+	s.BaseURL = "http://127.0.0.1:1"
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	out := make(chan recon.Finding, 1)
+	err := s.Sweep(ctx, "", out)
+	if !errors.Is(err, context.Canceled) {
+		t.Fatalf("expected context.Canceled, got %v", err)
+	}
+}
+
+func TestYandexSource_Unauthorized(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("yandex", 1000, 100)
+
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.WriteHeader(http.StatusUnauthorized)
+		_, _ = w.Write([]byte("bad creds"))
+	}))
+	defer srv.Close()
+
+	s := NewYandexSource("user", "key", reg, lim)
+	s.BaseURL = srv.URL
+
+	out := make(chan recon.Finding, 1)
+	err := s.Sweep(context.Background(), "", out)
+	if !errors.Is(err, ErrUnauthorized) {
+		t.Fatalf("expected ErrUnauthorized, got %v", err)
+	}
+}