feat(13-03): implement DockerHubSource and KubernetesSource

- DockerHub searches hub.docker.com v2 search API for repos matching provider keywords
- Kubernetes searches Artifact Hub for operators/manifests with kind-aware URL paths
- Both sources: context cancellation, nil registry, httptest-based tests

pkg/recon/sources/dockerhub.go

@@ -0,0 +1,125 @@
+package sources
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"net/http"
+	"net/url"
+	"time"
+
+	"golang.org/x/time/rate"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+// DockerHubSource searches Docker Hub for public images matching provider
+// keywords. Unauthenticated search is rate-limited but freely accessible.
+//
+// Emits one Finding per repository result, tagged SourceType=recon:dockerhub.
+type DockerHubSource struct {
+	// BaseURL defaults to https://hub.docker.com. Tests override with httptest URL.
+	BaseURL string
+	// Registry drives the keyword query list via BuildQueries.
+	Registry *providers.Registry
+	// Limiters is the shared recon.LimiterRegistry.
+	Limiters *recon.LimiterRegistry
+	// Client is the shared retry HTTP wrapper. If nil, a default is used.
+	Client *Client
+}
+
+// Compile-time assertion that DockerHubSource satisfies recon.ReconSource.
+var _ recon.ReconSource = (*DockerHubSource)(nil)
+
+func (s *DockerHubSource) Name() string          { return "dockerhub" }
+func (s *DockerHubSource) RateLimit() rate.Limit { return rate.Every(2 * time.Second) }
+func (s *DockerHubSource) Burst() int            { return 2 }
+func (s *DockerHubSource) RespectsRobots() bool  { return false }
+
+// Enabled always returns true: Docker Hub search is unauthenticated.
+func (s *DockerHubSource) Enabled(_ recon.Config) bool { return true }
+
+// Sweep iterates provider keywords, searches Docker Hub for matching
+// repositories, and emits a Finding for each result.
+func (s *DockerHubSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
+	base := s.BaseURL
+	if base == "" {
+		base = "https://hub.docker.com"
+	}
+	client := s.Client
+	if client == nil {
+		client = NewClient()
+	}
+
+	queries := BuildQueries(s.Registry, "dockerhub")
+	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
+			}
+		}
+
+		endpoint := fmt.Sprintf("%s/v2/search/repositories/?query=%s&page_size=20",
+			base, url.QueryEscape(q))
+		req, err := http.NewRequestWithContext(ctx, http.MethodGet, endpoint, nil)
+		if err != nil {
+			return fmt.Errorf("dockerhub: build req: %w", err)
+		}
+		req.Header.Set("Accept", "application/json")
+
+		resp, err := client.Do(ctx, req)
+		if err != nil {
+			// Non-fatal: skip this keyword on transient errors.
+			continue
+		}
+
+		var parsed dockerHubSearchResponse
+		decErr := json.NewDecoder(resp.Body).Decode(&parsed)
+		_ = resp.Body.Close()
+		if decErr != nil {
+			continue
+		}
+
+		for _, repo := range parsed.Results {
+			if err := ctx.Err(); err != nil {
+				return err
+			}
+			sourceURL := fmt.Sprintf("https://hub.docker.com/r/%s", repo.RepoName)
+			if base != "https://hub.docker.com" {
+				sourceURL = fmt.Sprintf("%s/r/%s", base, repo.RepoName)
+			}
+			f := recon.Finding{
+				ProviderName: "",
+				Source:       sourceURL,
+				SourceType:   "recon:dockerhub",
+				Confidence:   "low",
+				DetectedAt:   time.Now(),
+			}
+			select {
+			case out <- f:
+			case <-ctx.Done():
+				return ctx.Err()
+			}
+		}
+	}
+	return nil
+}
+
+type dockerHubSearchResponse struct {
+	Results []dockerHubRepo `json:"results"`
+}
+
+type dockerHubRepo struct {
+	RepoName    string `json:"repo_name"`
+	Description string `json:"description"`
+	IsOfficial  bool   `json:"is_official"`
+}

pkg/recon/sources/dockerhub_test.go

@@ -0,0 +1,130 @@
+package sources
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"net/http"
+	"net/http/httptest"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+func dockerHubStubHandler(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 != "/v2/search/repositories/" {
+			t.Errorf("unexpected path: %s", r.URL.Path)
+		}
+		if r.URL.Query().Get("query") == "" {
+			t.Errorf("missing query param")
+		}
+		body := dockerHubSearchResponse{
+			Results: []dockerHubRepo{
+				{RepoName: "alice/openai-proxy", Description: "OpenAI proxy", IsOfficial: false},
+				{RepoName: "bob/llm-gateway", Description: "LLM gateway", IsOfficial: false},
+			},
+		}
+		w.Header().Set("Content-Type", "application/json")
+		_ = json.NewEncoder(w).Encode(body)
+	}
+}
+
+func TestDockerHub_SweepEmitsFindings(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("dockerhub", 1000, 100)
+
+	var calls int32
+	srv := httptest.NewServer(dockerHubStubHandler(t, &calls))
+	defer srv.Close()
+
+	src := &DockerHubSource{
+		BaseURL:  srv.URL,
+		Registry: reg,
+		Limiters: lim,
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 32)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	done := make(chan error, 1)
+	go func() { done <- src.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 results = 4 findings
+	if len(findings) != 4 {
+		t.Fatalf("expected 4 findings, got %d", len(findings))
+	}
+	for _, f := range findings {
+		if f.SourceType != "recon:dockerhub" {
+			t.Errorf("SourceType=%q want recon:dockerhub", f.SourceType)
+		}
+	}
+	if got := atomic.LoadInt32(&calls); got != 2 {
+		t.Errorf("expected 2 server calls, got %d", got)
+	}
+}
+
+func TestDockerHub_EnabledAlwaysTrue(t *testing.T) {
+	s := &DockerHubSource{}
+	if !s.Enabled(recon.Config{}) {
+		t.Fatal("expected Enabled=true")
+	}
+}
+
+func TestDockerHub_NameAndRate(t *testing.T) {
+	s := &DockerHubSource{}
+	if s.Name() != "dockerhub" {
+		t.Errorf("unexpected name: %s", s.Name())
+	}
+	if s.Burst() != 2 {
+		t.Errorf("burst: %d", s.Burst())
+	}
+	if s.RespectsRobots() {
+		t.Error("expected RespectsRobots=false")
+	}
+}
+
+func TestDockerHub_CtxCancelled(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("dockerhub", 1000, 100)
+
+	src := &DockerHubSource{
+		BaseURL:  "http://127.0.0.1:1",
+		Registry: reg,
+		Limiters: lim,
+		Client:   NewClient(),
+	}
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	out := make(chan recon.Finding, 1)
+	err := src.Sweep(ctx, "", out)
+	if !errors.Is(err, context.Canceled) {
+		t.Fatalf("expected context.Canceled, got %v", err)
+	}
+}
+
+func TestDockerHub_NilRegistryNoError(t *testing.T) {
+	src := &DockerHubSource{Client: NewClient()}
+	out := make(chan recon.Finding, 1)
+	if err := src.Sweep(context.Background(), "", out); err != nil {
+		t.Fatalf("expected nil, got %v", err)
+	}
+}

pkg/recon/sources/kubernetes.go

@@ -0,0 +1,156 @@
+package sources
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"net/http"
+	"net/url"
+	"time"
+
+	"golang.org/x/time/rate"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+// KubernetesSource searches Artifact Hub for Kubernetes operators and manifests
+// matching provider keywords. This discovers publicly published K8s packages
+// that may embed API keys in their manifests or values files.
+//
+// Emits one Finding per package result, tagged SourceType=recon:k8s.
+type KubernetesSource struct {
+	// BaseURL defaults to https://artifacthub.io. Tests override with httptest URL.
+	BaseURL string
+	// Registry drives the keyword query list via BuildQueries.
+	Registry *providers.Registry
+	// Limiters is the shared recon.LimiterRegistry.
+	Limiters *recon.LimiterRegistry
+	// Client is the shared retry HTTP wrapper. If nil, a default is used.
+	Client *Client
+}
+
+// Compile-time assertion that KubernetesSource satisfies recon.ReconSource.
+var _ recon.ReconSource = (*KubernetesSource)(nil)
+
+func (s *KubernetesSource) Name() string          { return "k8s" }
+func (s *KubernetesSource) RateLimit() rate.Limit { return rate.Every(3 * time.Second) }
+func (s *KubernetesSource) Burst() int            { return 1 }
+func (s *KubernetesSource) RespectsRobots() bool  { return true }
+
+// Enabled always returns true: Artifact Hub search is unauthenticated.
+func (s *KubernetesSource) Enabled(_ recon.Config) bool { return true }
+
+// Sweep iterates provider keywords, searches Artifact Hub for Kubernetes
+// operators (kind=6), and emits a Finding for each result.
+func (s *KubernetesSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
+	base := s.BaseURL
+	if base == "" {
+		base = "https://artifacthub.io"
+	}
+	client := s.Client
+	if client == nil {
+		client = NewClient()
+	}
+
+	queries := BuildQueries(s.Registry, "k8s")
+	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
+			}
+		}
+
+		// kind left empty to search across all Kubernetes-related package types.
+		endpoint := fmt.Sprintf("%s/api/v1/packages/search?ts_query_web=%s&limit=20",
+			base, url.QueryEscape(q))
+		req, err := http.NewRequestWithContext(ctx, http.MethodGet, endpoint, nil)
+		if err != nil {
+			return fmt.Errorf("k8s: build req: %w", err)
+		}
+		req.Header.Set("Accept", "application/json")
+
+		resp, err := client.Do(ctx, req)
+		if err != nil {
+			continue
+		}
+
+		var parsed k8sSearchResponse
+		decErr := json.NewDecoder(resp.Body).Decode(&parsed)
+		_ = resp.Body.Close()
+		if decErr != nil {
+			continue
+		}
+
+		for _, pkg := range parsed.Packages {
+			if err := ctx.Err(); err != nil {
+				return err
+			}
+
+			repoName := ""
+			if pkg.Repository.Name != "" {
+				repoName = pkg.Repository.Name
+			}
+
+			kindPath := k8sKindPath(pkg.Repository.Kind)
+			sourceURL := fmt.Sprintf("https://artifacthub.io/packages/%s/%s/%s",
+				kindPath, repoName, pkg.NormalizedName)
+			if base != "https://artifacthub.io" {
+				sourceURL = fmt.Sprintf("%s/packages/%s/%s/%s",
+					base, kindPath, repoName, pkg.NormalizedName)
+			}
+
+			f := recon.Finding{
+				ProviderName: "",
+				Source:       sourceURL,
+				SourceType:   "recon:k8s",
+				Confidence:   "low",
+				DetectedAt:   time.Now(),
+			}
+			select {
+			case out <- f:
+			case <-ctx.Done():
+				return ctx.Err()
+			}
+		}
+	}
+	return nil
+}
+
+type k8sSearchResponse struct {
+	Packages []k8sPackage `json:"packages"`
+}
+
+type k8sPackage struct {
+	PackageID      string    `json:"package_id"`
+	Name           string    `json:"name"`
+	NormalizedName string    `json:"normalized_name"`
+	Repository     k8sRepo   `json:"repository"`
+}
+
+type k8sRepo struct {
+	Name string `json:"name"`
+	Kind int    `json:"kind"`
+}
+
+// k8sKindPath maps Artifact Hub kind integers to URL path segments.
+func k8sKindPath(kind int) string {
+	switch kind {
+	case 0:
+		return "helm"
+	case 6:
+		return "kube-operator"
+	case 7:
+		return "kubectl"
+	default:
+		return "other"
+	}
+}

pkg/recon/sources/kubernetes_test.go

@@ -0,0 +1,200 @@
+package sources
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"net/http"
+	"net/http/httptest"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+func k8sStubHandler(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 != "/api/v1/packages/search" {
+			t.Errorf("unexpected path: %s", r.URL.Path)
+		}
+		if r.URL.Query().Get("ts_query_web") == "" {
+			t.Errorf("missing ts_query_web param")
+		}
+		body := k8sSearchResponse{
+			Packages: []k8sPackage{
+				{
+					PackageID:      "pkg-1",
+					Name:           "openai-operator",
+					NormalizedName: "openai-operator",
+					Repository:     k8sRepo{Name: "community", Kind: 6},
+				},
+				{
+					PackageID:      "pkg-2",
+					Name:           "llm-secrets",
+					NormalizedName: "llm-secrets",
+					Repository:     k8sRepo{Name: "stable", Kind: 0},
+				},
+			},
+		}
+		w.Header().Set("Content-Type", "application/json")
+		_ = json.NewEncoder(w).Encode(body)
+	}
+}
+
+func TestKubernetes_SweepEmitsFindings(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("k8s", 1000, 100)
+
+	var calls int32
+	srv := httptest.NewServer(k8sStubHandler(t, &calls))
+	defer srv.Close()
+
+	src := &KubernetesSource{
+		BaseURL:  srv.URL,
+		Registry: reg,
+		Limiters: lim,
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 32)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	done := make(chan error, 1)
+	go func() { done <- src.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 results = 4 findings
+	if len(findings) != 4 {
+		t.Fatalf("expected 4 findings, got %d", len(findings))
+	}
+	for _, f := range findings {
+		if f.SourceType != "recon:k8s" {
+			t.Errorf("SourceType=%q want recon:k8s", f.SourceType)
+		}
+	}
+	if got := atomic.LoadInt32(&calls); got != 2 {
+		t.Errorf("expected 2 server calls, got %d", got)
+	}
+}
+
+func TestKubernetes_KindPaths(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("k8s", 1000, 100)
+
+	var calls int32
+	srv := httptest.NewServer(k8sStubHandler(t, &calls))
+	defer srv.Close()
+
+	src := &KubernetesSource{
+		BaseURL:  srv.URL,
+		Registry: reg,
+		Limiters: lim,
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 32)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	done := make(chan error, 1)
+	go func() { done <- src.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)
+	}
+
+	// Check that kind=6 maps to kube-operator and kind=0 maps to helm in URLs.
+	hasOperator := false
+	hasHelm := false
+	for _, f := range findings {
+		if contains(f.Source, "/kube-operator/") {
+			hasOperator = true
+		}
+		if contains(f.Source, "/helm/") {
+			hasHelm = true
+		}
+	}
+	if !hasOperator {
+		t.Error("expected at least one finding with kube-operator path")
+	}
+	if !hasHelm {
+		t.Error("expected at least one finding with helm path")
+	}
+}
+
+func TestKubernetes_EnabledAlwaysTrue(t *testing.T) {
+	s := &KubernetesSource{}
+	if !s.Enabled(recon.Config{}) {
+		t.Fatal("expected Enabled=true")
+	}
+}
+
+func TestKubernetes_NameAndRate(t *testing.T) {
+	s := &KubernetesSource{}
+	if s.Name() != "k8s" {
+		t.Errorf("unexpected name: %s", s.Name())
+	}
+	if s.Burst() != 1 {
+		t.Errorf("burst: %d", s.Burst())
+	}
+	if !s.RespectsRobots() {
+		t.Error("expected RespectsRobots=true")
+	}
+}
+
+func TestKubernetes_CtxCancelled(t *testing.T) {
+	reg := syntheticRegistry()
+	lim := recon.NewLimiterRegistry()
+	_ = lim.For("k8s", 1000, 100)
+
+	src := &KubernetesSource{
+		BaseURL:  "http://127.0.0.1:1",
+		Registry: reg,
+		Limiters: lim,
+		Client:   NewClient(),
+	}
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	out := make(chan recon.Finding, 1)
+	err := src.Sweep(ctx, "", out)
+	if !errors.Is(err, context.Canceled) {
+		t.Fatalf("expected context.Canceled, got %v", err)
+	}
+}
+
+func TestKubernetes_NilRegistryNoError(t *testing.T) {
+	src := &KubernetesSource{Client: NewClient()}
+	out := make(chan recon.Finding, 1)
+	if err := src.Sweep(context.Background(), "", out); err != nil {
+		t.Fatalf("expected nil, got %v", err)
+	}
+}
+
+// contains checks if substr is in s. Avoids importing strings in test.
+func contains(s, substr string) bool {
+	for i := 0; i+len(substr) <= len(s); i++ {
+		if s[i:i+len(substr)] == substr {
+			return true
+		}
+	}
+	return false
+}