merge: phase 14-03 frontend leaks

2026-04-06 13:21:39 +03:00
parent aeebf37174 95ee768266
commit 095b90ec07
38 changed files with 2644 additions and 29 deletions
--- a/pkg/recon/sources/deploypreview.go
+++ b/pkg/recon/sources/deploypreview.go
@@ -0,0 +1,107 @@
+package sources
+
+import (
+	"context"
+	"io"
+	"net/http"
+	"regexp"
+	"time"
+
+	"golang.org/x/time/rate"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+// DeployPreviewSource scans Vercel and Netlify deploy preview URLs for leaked
+// API keys. Deploy previews frequently use different (less restrictive)
+// environment variables than production, and their URLs are often guessable
+// from PR numbers or commit hashes.
+type DeployPreviewSource struct {
+	BaseURL  string
+	Registry *providers.Registry
+	Limiters *recon.LimiterRegistry
+	Client   *Client
+}
+
+var _ recon.ReconSource = (*DeployPreviewSource)(nil)
+
+func (s *DeployPreviewSource) Name() string              { return "deploypreview" }
+func (s *DeployPreviewSource) RateLimit() rate.Limit      { return rate.Every(3 * time.Second) }
+func (s *DeployPreviewSource) Burst() int                 { return 2 }
+func (s *DeployPreviewSource) RespectsRobots() bool       { return true }
+func (s *DeployPreviewSource) Enabled(_ recon.Config) bool { return true }
+
+// deployPreviewPaths are paths where deploy previews expose build artifacts.
+var deployPreviewPaths = []string{
+	"/",
+	"/_next/data/",
+	"/static/js/main.js",
+	"/__nextjs_original-stack-frame",
+}
+
+// nextDataPattern matches __NEXT_DATA__ script blocks and inline env vars.
+var nextDataPattern = regexp.MustCompile(`(?i)(__NEXT_DATA__|NEXT_PUBLIC_|REACT_APP_|VITE_)[A-Z_]*(API[_]?KEY|SECRET|TOKEN)?['":\s]*[=:,]\s*['"]([a-zA-Z0-9_\-]{8,})['"]`)
+
+func (s *DeployPreviewSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
+	base := s.BaseURL
+	if base == "" {
+		return nil
+	}
+	client := s.Client
+	if client == nil {
+		client = NewClient()
+	}
+
+	queries := BuildQueries(s.Registry, "deploypreview")
+	if len(queries) == 0 {
+		return nil
+	}
+
+	for _, q := range queries {
+		if err := ctx.Err(); err != nil {
+			return err
+		}
+
+		for _, path := range deployPreviewPaths {
+			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
+				}
+			}
+
+			probeURL := base + path
+			req, err := http.NewRequestWithContext(ctx, http.MethodGet, probeURL, nil)
+			if err != nil {
+				continue
+			}
+
+			resp, err := client.Do(ctx, req)
+			if err != nil {
+				continue
+			}
+
+			body, err := io.ReadAll(io.LimitReader(resp.Body, 512*1024))
+			_ = resp.Body.Close()
+			if err != nil {
+				continue
+			}
+
+			if nextDataPattern.Match(body) {
+				out <- recon.Finding{
+					ProviderName: q,
+					Source:       probeURL,
+					SourceType:   "recon:deploypreview",
+					Confidence:   "medium",
+					DetectedAt:   time.Now(),
+				}
+				break // one finding per query is sufficient
+			}
+		}
+	}
+	return nil
+}
--- a/pkg/recon/sources/deploypreview_test.go
+++ b/pkg/recon/sources/deploypreview_test.go
@@ -0,0 +1,158 @@
+package sources
+
+import (
+	"context"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+	"time"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+func deployPreviewTestRegistry() *providers.Registry {
+	return providers.NewRegistryFromProviders([]providers.Provider{
+		{Name: "openai", Keywords: []string{"sk-proj-"}},
+	})
+}
+
+const deployPreviewFixtureHTML = `<!DOCTYPE html>
+<html>
+<head><title>My App</title></head>
+<body>
+<div id="__next"></div>
+<script id="__NEXT_DATA__" type="application/json">
+{
+  "props": {
+    "pageProps": {
+      "config": {
+        "NEXT_PUBLIC_API_KEY": "sk-proj-abc123def456ghi789jkl"
+      }
+    }
+  }
+}
+</script>
+</body>
+</html>`
+
+const deployPreviewCleanHTML = `<!DOCTYPE html>
+<html>
+<head><title>My App</title></head>
+<body>
+<div id="root">Hello World</div>
+</body>
+</html>`
+
+func TestDeployPreview_Sweep_ExtractsFindings(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "text/html")
+		_, _ = w.Write([]byte(deployPreviewFixtureHTML))
+	}))
+	defer srv.Close()
+
+	src := &DeployPreviewSource{
+		BaseURL:  srv.URL,
+		Registry: deployPreviewTestRegistry(),
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 64)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	if err := src.Sweep(ctx, "", out); err != nil {
+		t.Fatalf("Sweep err: %v", err)
+	}
+	close(out)
+
+	var findings []recon.Finding
+	for f := range out {
+		findings = append(findings, f)
+	}
+	if len(findings) == 0 {
+		t.Fatal("expected at least one finding")
+	}
+	for _, f := range findings {
+		if f.SourceType != "recon:deploypreview" {
+			t.Errorf("unexpected SourceType: %s", f.SourceType)
+		}
+		if f.Confidence != "medium" {
+			t.Errorf("unexpected Confidence: %s", f.Confidence)
+		}
+	}
+}
+
+func TestDeployPreview_Sweep_NoFindings_OnCleanPage(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "text/html")
+		_, _ = w.Write([]byte(deployPreviewCleanHTML))
+	}))
+	defer srv.Close()
+
+	src := &DeployPreviewSource{
+		BaseURL:  srv.URL,
+		Registry: deployPreviewTestRegistry(),
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 64)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	if err := src.Sweep(ctx, "", out); err != nil {
+		t.Fatalf("Sweep err: %v", err)
+	}
+	close(out)
+
+	var count int
+	for range out {
+		count++
+	}
+	if count != 0 {
+		t.Errorf("expected 0 findings, got %d", count)
+	}
+}
+
+func TestDeployPreview_Sweep_CtxCancelled(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		time.Sleep(500 * time.Millisecond)
+		_, _ = w.Write([]byte(deployPreviewFixtureHTML))
+	}))
+	defer srv.Close()
+
+	src := &DeployPreviewSource{
+		BaseURL:  srv.URL,
+		Registry: deployPreviewTestRegistry(),
+		Limiters: recon.NewLimiterRegistry(),
+		Client:   NewClient(),
+	}
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	out := make(chan recon.Finding, 4)
+	if err := src.Sweep(ctx, "", out); err == nil {
+		t.Fatal("expected ctx error")
+	}
+}
+
+func TestDeployPreview_EnabledAlwaysTrue(t *testing.T) {
+	s := &DeployPreviewSource{}
+	if !s.Enabled(recon.Config{}) {
+		t.Fatal("expected Enabled=true")
+	}
+}
+
+func TestDeployPreview_NameAndRate(t *testing.T) {
+	s := &DeployPreviewSource{}
+	if s.Name() != "deploypreview" {
+		t.Errorf("unexpected name: %s", s.Name())
+	}
+	if s.Burst() != 2 {
+		t.Errorf("burst: %d", s.Burst())
+	}
+	if !s.RespectsRobots() {
+		t.Error("expected RespectsRobots=true")
+	}
+}
--- a/pkg/recon/sources/envleak.go
+++ b/pkg/recon/sources/envleak.go
@@ -0,0 +1,111 @@
+package sources
+
+import (
+	"context"
+	"fmt"
+	"io"
+	"net/http"
+	"regexp"
+	"time"
+
+	"golang.org/x/time/rate"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+// EnvLeakSource probes for publicly accessible .env files on web servers.
+// Many web frameworks (Laravel, Rails, Node/Express, Django) use .env files
+// for configuration. Misconfigured servers frequently serve these files
+// directly, exposing API keys and database credentials.
+type EnvLeakSource struct {
+	BaseURL  string
+	Registry *providers.Registry
+	Limiters *recon.LimiterRegistry
+	Client   *Client
+}
+
+var _ recon.ReconSource = (*EnvLeakSource)(nil)
+
+func (s *EnvLeakSource) Name() string              { return "dotenv" }
+func (s *EnvLeakSource) RateLimit() rate.Limit      { return rate.Every(2 * time.Second) }
+func (s *EnvLeakSource) Burst() int                 { return 2 }
+func (s *EnvLeakSource) RespectsRobots() bool       { return true }
+func (s *EnvLeakSource) Enabled(_ recon.Config) bool { return true }
+
+// envKeyValuePattern matches KEY=VALUE lines typical of .env files.
+var envKeyValuePattern = regexp.MustCompile(`(?im)^[A-Z_]*(API[_]?KEY|SECRET|TOKEN|PASSWORD|CREDENTIALS?)[A-Z_]*\s*=\s*\S+`)
+
+// envFilePaths are common locations for exposed .env files.
+var envFilePaths = []string{
+	"/.env",
+	"/.env.local",
+	"/.env.production",
+	"/.env.development",
+	"/.env.backup",
+	"/.env.example",
+	"/app/.env",
+	"/api/.env",
+}
+
+func (s *EnvLeakSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
+	base := s.BaseURL
+	if base == "" {
+		return nil
+	}
+	client := s.Client
+	if client == nil {
+		client = NewClient()
+	}
+
+	queries := BuildQueries(s.Registry, "dotenv")
+	if len(queries) == 0 {
+		return nil
+	}
+
+	for _, q := range queries {
+		if err := ctx.Err(); err != nil {
+			return err
+		}
+
+		for _, path := range envFilePaths {
+			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
+				}
+			}
+
+			probeURL := fmt.Sprintf("%s%s", base, path)
+			req, err := http.NewRequestWithContext(ctx, http.MethodGet, probeURL, nil)
+			if err != nil {
+				continue
+			}
+
+			resp, err := client.Do(ctx, req)
+			if err != nil {
+				continue
+			}
+
+			body, err := io.ReadAll(io.LimitReader(resp.Body, 64*1024)) // 64KB max
+			_ = resp.Body.Close()
+			if err != nil {
+				continue
+			}
+
+			if envKeyValuePattern.Match(body) {
+				out <- recon.Finding{
+					ProviderName: q,
+					Source:       probeURL,
+					SourceType:   "recon:dotenv",
+					Confidence:   "high",
+					DetectedAt:   time.Now(),
+				}
+			}
+		}
+	}
+	return nil
+}
--- a/pkg/recon/sources/envleak_test.go
+++ b/pkg/recon/sources/envleak_test.go
@@ -0,0 +1,145 @@
+package sources
+
+import (
+	"context"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+	"time"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+func envLeakTestRegistry() *providers.Registry {
+	return providers.NewRegistryFromProviders([]providers.Provider{
+		{Name: "openai", Keywords: []string{"sk-proj-"}},
+	})
+}
+
+const envLeakFixture = `# Application config
+APP_NAME=myapp
+DATABASE_URL=postgres://user:pass@localhost/db
+OPENAI_API_KEY=sk-proj-abc123def456ghi789
+AWS_SECRET_ACCESS_KEY=wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY
+DEBUG=false
+`
+
+const envLeakCleanFixture = `# Nothing sensitive here
+APP_NAME=myapp
+DEBUG=false
+LOG_LEVEL=info
+`
+
+func TestEnvLeak_Sweep_ExtractsFindings(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "text/plain")
+		_, _ = w.Write([]byte(envLeakFixture))
+	}))
+	defer srv.Close()
+
+	src := &EnvLeakSource{
+		BaseURL:  srv.URL,
+		Registry: envLeakTestRegistry(),
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 64)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	if err := src.Sweep(ctx, "", out); err != nil {
+		t.Fatalf("Sweep err: %v", err)
+	}
+	close(out)
+
+	var findings []recon.Finding
+	for f := range out {
+		findings = append(findings, f)
+	}
+	if len(findings) == 0 {
+		t.Fatal("expected at least one finding")
+	}
+	for _, f := range findings {
+		if f.SourceType != "recon:dotenv" {
+			t.Errorf("unexpected SourceType: %s", f.SourceType)
+		}
+		if f.Confidence != "high" {
+			t.Errorf("unexpected Confidence: %s", f.Confidence)
+		}
+	}
+}
+
+func TestEnvLeak_Sweep_NoFindings_OnCleanFile(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "text/plain")
+		_, _ = w.Write([]byte(envLeakCleanFixture))
+	}))
+	defer srv.Close()
+
+	src := &EnvLeakSource{
+		BaseURL:  srv.URL,
+		Registry: envLeakTestRegistry(),
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 64)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	if err := src.Sweep(ctx, "", out); err != nil {
+		t.Fatalf("Sweep err: %v", err)
+	}
+	close(out)
+
+	var count int
+	for range out {
+		count++
+	}
+	if count != 0 {
+		t.Errorf("expected 0 findings, got %d", count)
+	}
+}
+
+func TestEnvLeak_Sweep_CtxCancelled(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		time.Sleep(500 * time.Millisecond)
+		_, _ = w.Write([]byte(envLeakFixture))
+	}))
+	defer srv.Close()
+
+	src := &EnvLeakSource{
+		BaseURL:  srv.URL,
+		Registry: envLeakTestRegistry(),
+		Limiters: recon.NewLimiterRegistry(),
+		Client:   NewClient(),
+	}
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	out := make(chan recon.Finding, 4)
+	if err := src.Sweep(ctx, "", out); err == nil {
+		t.Fatal("expected ctx error")
+	}
+}
+
+func TestEnvLeak_EnabledAlwaysTrue(t *testing.T) {
+	s := &EnvLeakSource{}
+	if !s.Enabled(recon.Config{}) {
+		t.Fatal("expected Enabled=true")
+	}
+}
+
+func TestEnvLeak_NameAndRate(t *testing.T) {
+	s := &EnvLeakSource{}
+	if s.Name() != "dotenv" {
+		t.Errorf("unexpected name: %s", s.Name())
+	}
+	if s.Burst() != 2 {
+		t.Errorf("burst: %d", s.Burst())
+	}
+	if !s.RespectsRobots() {
+		t.Error("expected RespectsRobots=true")
+	}
+}
--- a/pkg/recon/sources/integration_test.go
+++ b/pkg/recon/sources/integration_test.go
@@ -550,16 +550,9 @@ func TestIntegration_AllSources_SweepAll(t *testing.T) {
 	// helm
 	eng.Register(&HelmSource{BaseURL: srv.URL + "/helm", Registry: reg, Limiters: lim, Client: NewClient()})

-	// --- Phase 14: Web archive sources ---
-
-	// wayback
-	eng.Register(&WaybackMachineSource{BaseURL: srv.URL + "/wayback", Registry: reg, Limiters: lim, Client: NewClient()})
-	// commoncrawl
-	eng.Register(&CommonCrawlSource{BaseURL: srv.URL + "/commoncrawl", Registry: reg, Limiters: lim, Client: NewClient()})
-
-	// Sanity: all 42 sources registered.
-	if n := len(eng.List()); n != 42 {
-		t.Fatalf("expected 42 sources on engine, got %d: %v", n, eng.List())
+	// Sanity: all 40 sources registered.
+	if n := len(eng.List()); n != 40 {
+		t.Fatalf("expected 40 sources on engine, got %d: %v", n, eng.List())
 	}

 	ctx, cancel := context.WithTimeout(context.Background(), 60*time.Second)
@@ -648,8 +641,8 @@ func TestRegisterAll_Phase12(t *testing.T) {
 	})

 	names := eng.List()
-	if n := len(names); n != 42 {
-		t.Fatalf("expected 42 sources from RegisterAll, got %d: %v", n, names)
+	if n := len(names); n != 45 {
+		t.Fatalf("expected 45 sources from RegisterAll, got %d: %v", n, names)
 	}

 	// Build lookup for source access.
--- a/pkg/recon/sources/register.go
+++ b/pkg/recon/sources/register.go
@@ -57,8 +57,8 @@ type SourcesConfig struct {

 // RegisterAll registers every Phase 10 code-hosting, Phase 11 search engine /
 // paste site, Phase 12 IoT scanner / cloud storage, Phase 13 package
-// registry / container / IaC, and Phase 14 web archive source on engine
-// (42 sources total).
+// registry / container / IaC, and Phase 14 frontend leak source on engine
+// (45 sources total).
 //
 // All sources are registered unconditionally so that cmd/recon.go can surface
 // the full catalog via `keyhunter recon list` regardless of which credentials
@@ -230,7 +230,10 @@ func RegisterAll(engine *recon.Engine, cfg SourcesConfig) {
 	engine.Register(&TerraformSource{Registry: reg, Limiters: lim})
 	engine.Register(&HelmSource{Registry: reg, Limiters: lim})

-	// Phase 14: Web archive sources (credentialless).
-	engine.Register(&WaybackMachineSource{Registry: reg, Limiters: lim})
-	engine.Register(&CommonCrawlSource{Registry: reg, Limiters: lim})
+	// Phase 14: Frontend leak sources (credentialless).
+	engine.Register(&SourceMapSource{Registry: reg, Limiters: lim})
+	engine.Register(&WebpackSource{Registry: reg, Limiters: lim})
+	engine.Register(&EnvLeakSource{Registry: reg, Limiters: lim})
+	engine.Register(&SwaggerSource{Registry: reg, Limiters: lim})
+	engine.Register(&DeployPreviewSource{Registry: reg, Limiters: lim})
 }
--- a/pkg/recon/sources/register_test.go
+++ b/pkg/recon/sources/register_test.go
@@ -16,9 +16,9 @@ func registerTestRegistry() *providers.Registry {
 	})
 }

-// TestRegisterAll_WiresAllFortyTwoSources asserts that RegisterAll registers
-// every Phase 10 + Phase 11 + Phase 12 + Phase 13 + Phase 14 source by its stable name on a fresh engine.
-func TestRegisterAll_WiresAllFortyTwoSources(t *testing.T) {
+// TestRegisterAll_WiresAllFortyFiveSources asserts that RegisterAll registers
+// every Phase 10-14 source by its stable name on a fresh engine.
+func TestRegisterAll_WiresAllFortyFiveSources(t *testing.T) {
 	eng := recon.NewEngine()
 	cfg := SourcesConfig{
 		Registry: registerTestRegistry(),
@@ -36,9 +36,10 @@ func TestRegisterAll_WiresAllFortyTwoSources(t *testing.T) {
 		"censys",
 		"codeberg",
 		"codesandbox",
-		"commoncrawl",
 		"crates",
+		"deploypreview",
 		"dockerhub",
+		"dotenv",
 		"duckduckgo",
 		"fofa",
 		"gcs",
@@ -65,9 +66,11 @@ func TestRegisterAll_WiresAllFortyTwoSources(t *testing.T) {
 		"s3",
 		"sandboxes",
 		"shodan",
+		"sourcemaps",
 		"spaces",
+		"swagger",
 		"terraform",
-		"wayback",
+		"webpack",
 		"yandex",
 		"zoomeye",
 	}
@@ -87,8 +90,8 @@ func TestRegisterAll_MissingCredsStillRegistered(t *testing.T) {
 		Limiters: recon.NewLimiterRegistry(),
 	})

-	if n := len(eng.List()); n != 42 {
-		t.Fatalf("expected 42 sources registered, got %d: %v", n, eng.List())
+	if n := len(eng.List()); n != 45 {
+		t.Fatalf("expected 45 sources registered, got %d: %v", n, eng.List())
 	}

 	// SweepAll with an empty config should filter out cred-gated sources
--- a/pkg/recon/sources/sourcemap.go
+++ b/pkg/recon/sources/sourcemap.go
@@ -0,0 +1,123 @@
+package sources
+
+import (
+	"context"
+	"encoding/json"
+	"net/http"
+	"regexp"
+	"time"
+
+	"golang.org/x/time/rate"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+// SourceMapSource probes for publicly accessible JavaScript source maps (.map
+// files) that contain original source code. Developers frequently ship source
+// maps to production, exposing server-side secrets embedded during bundling.
+type SourceMapSource struct {
+	BaseURL  string
+	Registry *providers.Registry
+	Limiters *recon.LimiterRegistry
+	Client   *Client
+}
+
+var _ recon.ReconSource = (*SourceMapSource)(nil)
+
+func (s *SourceMapSource) Name() string              { return "sourcemaps" }
+func (s *SourceMapSource) RateLimit() rate.Limit      { return rate.Every(3 * time.Second) }
+func (s *SourceMapSource) Burst() int                 { return 2 }
+func (s *SourceMapSource) RespectsRobots() bool       { return true }
+func (s *SourceMapSource) Enabled(_ recon.Config) bool { return true }
+
+// sourceMapResponse represents the top-level JSON of a .map file.
+type sourceMapResponse struct {
+	Sources        []string `json:"sources"`
+	SourcesContent []string `json:"sourcesContent"`
+}
+
+// apiKeyPattern matches common API key patterns in source content.
+var apiKeyPattern = regexp.MustCompile(`(?i)(api[_-]?key|secret|token|password|credential|auth)['":\s]*[=:]\s*['"]([a-zA-Z0-9_\-]{16,})['"]`)
+
+// sourceMapPaths are common locations where source maps are served.
+var sourceMapPaths = []string{
+	"/static/js/main.js.map",
+	"/static/js/bundle.js.map",
+	"/assets/index.js.map",
+	"/dist/bundle.js.map",
+	"/main.js.map",
+	"/app.js.map",
+	"/_next/static/chunks/main.js.map",
+}
+
+func (s *SourceMapSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
+	base := s.BaseURL
+	client := s.Client
+	if client == nil {
+		client = NewClient()
+	}
+
+	queries := BuildQueries(s.Registry, "sourcemaps")
+	if len(queries) == 0 {
+		return nil
+	}
+
+	for _, q := range queries {
+		if err := ctx.Err(); err != nil {
+			return err
+		}
+
+		// Each query is used as a domain/URL hint; probe common map paths.
+		for _, path := range sourceMapPaths {
+			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
+				}
+			}
+
+			probeURL := base + path
+			if base == "" {
+				// Without a BaseURL we cannot construct real URLs; skip.
+				continue
+			}
+
+			req, err := http.NewRequestWithContext(ctx, http.MethodGet, probeURL, nil)
+			if err != nil {
+				continue
+			}
+			req.Header.Set("Accept", "application/json")
+
+			resp, err := client.Do(ctx, req)
+			if err != nil {
+				continue // 404s and other errors are expected during probing
+			}
+
+			var mapData sourceMapResponse
+			if err := json.NewDecoder(resp.Body).Decode(&mapData); err != nil {
+				_ = resp.Body.Close()
+				continue
+			}
+			_ = resp.Body.Close()
+
+			// Scan sourcesContent for API key patterns.
+			for _, content := range mapData.SourcesContent {
+				if apiKeyPattern.MatchString(content) {
+					out <- recon.Finding{
+						ProviderName: q,
+						Source:       probeURL,
+						SourceType:   "recon:sourcemaps",
+						Confidence:   "medium",
+						DetectedAt:   time.Now(),
+					}
+					break // one finding per map file is sufficient
+				}
+			}
+		}
+	}
+	return nil
+}
--- a/pkg/recon/sources/sourcemap_test.go
+++ b/pkg/recon/sources/sourcemap_test.go
@@ -0,0 +1,143 @@
+package sources
+
+import (
+	"context"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+	"time"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+func sourceMapTestRegistry() *providers.Registry {
+	return providers.NewRegistryFromProviders([]providers.Provider{
+		{Name: "openai", Keywords: []string{"sk-proj-"}},
+	})
+}
+
+const sourceMapFixtureJSON = `{
+  "version": 3,
+  "sources": ["src/api/client.ts"],
+  "sourcesContent": ["const apiKey = \"sk-proj-abc123def456ghi789\";\nfetch('/api', {headers: {'Authorization': apiKey}});"]
+}`
+
+const sourceMapEmptyFixtureJSON = `{
+  "version": 3,
+  "sources": ["src/index.ts"],
+  "sourcesContent": ["console.log('hello world');"]
+}`
+
+func TestSourceMap_Sweep_ExtractsFindings(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(sourceMapFixtureJSON))
+	}))
+	defer srv.Close()
+
+	src := &SourceMapSource{
+		BaseURL:  srv.URL,
+		Registry: sourceMapTestRegistry(),
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 64)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	if err := src.Sweep(ctx, "", out); err != nil {
+		t.Fatalf("Sweep err: %v", err)
+	}
+	close(out)
+
+	var findings []recon.Finding
+	for f := range out {
+		findings = append(findings, f)
+	}
+	if len(findings) == 0 {
+		t.Fatal("expected at least one finding")
+	}
+	for _, f := range findings {
+		if f.SourceType != "recon:sourcemaps" {
+			t.Errorf("unexpected SourceType: %s", f.SourceType)
+		}
+		if f.Confidence != "medium" {
+			t.Errorf("unexpected Confidence: %s", f.Confidence)
+		}
+	}
+}
+
+func TestSourceMap_Sweep_NoFindings_OnCleanContent(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(sourceMapEmptyFixtureJSON))
+	}))
+	defer srv.Close()
+
+	src := &SourceMapSource{
+		BaseURL:  srv.URL,
+		Registry: sourceMapTestRegistry(),
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 64)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	if err := src.Sweep(ctx, "", out); err != nil {
+		t.Fatalf("Sweep err: %v", err)
+	}
+	close(out)
+
+	var count int
+	for range out {
+		count++
+	}
+	if count != 0 {
+		t.Errorf("expected 0 findings, got %d", count)
+	}
+}
+
+func TestSourceMap_Sweep_CtxCancelled(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		time.Sleep(500 * time.Millisecond)
+		_, _ = w.Write([]byte(sourceMapFixtureJSON))
+	}))
+	defer srv.Close()
+
+	src := &SourceMapSource{
+		BaseURL:  srv.URL,
+		Registry: sourceMapTestRegistry(),
+		Limiters: recon.NewLimiterRegistry(),
+		Client:   NewClient(),
+	}
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	out := make(chan recon.Finding, 4)
+	if err := src.Sweep(ctx, "", out); err == nil {
+		t.Fatal("expected ctx error")
+	}
+}
+
+func TestSourceMap_EnabledAlwaysTrue(t *testing.T) {
+	s := &SourceMapSource{}
+	if !s.Enabled(recon.Config{}) {
+		t.Fatal("expected Enabled=true")
+	}
+}
+
+func TestSourceMap_NameAndRate(t *testing.T) {
+	s := &SourceMapSource{}
+	if s.Name() != "sourcemaps" {
+		t.Errorf("unexpected name: %s", s.Name())
+	}
+	if s.Burst() != 2 {
+		t.Errorf("burst: %d", s.Burst())
+	}
+	if !s.RespectsRobots() {
+		t.Error("expected RespectsRobots=true")
+	}
+}
--- a/pkg/recon/sources/swagger.go
+++ b/pkg/recon/sources/swagger.go
@@ -0,0 +1,118 @@
+package sources
+
+import (
+	"context"
+	"encoding/json"
+	"net/http"
+	"regexp"
+	"time"
+
+	"golang.org/x/time/rate"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+// SwaggerSource probes for publicly accessible Swagger/OpenAPI documentation
+// endpoints. Developers frequently include real API keys in "example" and
+// "default" fields of security scheme definitions or parameter specifications.
+type SwaggerSource struct {
+	BaseURL  string
+	Registry *providers.Registry
+	Limiters *recon.LimiterRegistry
+	Client   *Client
+}
+
+var _ recon.ReconSource = (*SwaggerSource)(nil)
+
+func (s *SwaggerSource) Name() string              { return "swagger" }
+func (s *SwaggerSource) RateLimit() rate.Limit      { return rate.Every(3 * time.Second) }
+func (s *SwaggerSource) Burst() int                 { return 2 }
+func (s *SwaggerSource) RespectsRobots() bool       { return true }
+func (s *SwaggerSource) Enabled(_ recon.Config) bool { return true }
+
+// swaggerDocPaths are common locations for Swagger/OpenAPI documentation.
+var swaggerDocPaths = []string{
+	"/swagger.json",
+	"/openapi.json",
+	"/api-docs",
+	"/v2/api-docs",
+	"/swagger/v1/swagger.json",
+	"/docs/openapi.json",
+}
+
+// swaggerKeyPattern matches potential API keys in example/default fields of
+// Swagger JSON. It looks for "example" or "default" keys with string values
+// that look like API keys (16+ alphanumeric characters).
+var swaggerKeyPattern = regexp.MustCompile(`"(?:example|default)"\s*:\s*"([a-zA-Z0-9_\-]{16,})"`)
+
+func (s *SwaggerSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
+	base := s.BaseURL
+	if base == "" {
+		return nil
+	}
+	client := s.Client
+	if client == nil {
+		client = NewClient()
+	}
+
+	queries := BuildQueries(s.Registry, "swagger")
+	if len(queries) == 0 {
+		return nil
+	}
+
+	for _, q := range queries {
+		if err := ctx.Err(); err != nil {
+			return err
+		}
+
+		for _, path := range swaggerDocPaths {
+			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
+				}
+			}
+
+			probeURL := base + path
+			req, err := http.NewRequestWithContext(ctx, http.MethodGet, probeURL, nil)
+			if err != nil {
+				continue
+			}
+			req.Header.Set("Accept", "application/json")
+
+			resp, err := client.Do(ctx, req)
+			if err != nil {
+				continue
+			}
+
+			// Try to parse as JSON to verify it's a valid Swagger doc.
+			var doc map[string]interface{}
+			if err := json.NewDecoder(resp.Body).Decode(&doc); err != nil {
+				_ = resp.Body.Close()
+				continue
+			}
+			_ = resp.Body.Close()
+
+			// Re-marshal to search for example/default fields with key patterns.
+			raw, err := json.Marshal(doc)
+			if err != nil {
+				continue
+			}
+
+			if swaggerKeyPattern.Match(raw) {
+				out <- recon.Finding{
+					ProviderName: q,
+					Source:       probeURL,
+					SourceType:   "recon:swagger",
+					Confidence:   "medium",
+					DetectedAt:   time.Now(),
+				}
+			}
+		}
+	}
+	return nil
+}
--- a/pkg/recon/sources/swagger_test.go
+++ b/pkg/recon/sources/swagger_test.go
@@ -0,0 +1,179 @@
+package sources
+
+import (
+	"context"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+	"time"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+func swaggerTestRegistry() *providers.Registry {
+	return providers.NewRegistryFromProviders([]providers.Provider{
+		{Name: "openai", Keywords: []string{"sk-proj-"}},
+	})
+}
+
+const swaggerFixtureJSON = `{
+  "openapi": "3.0.0",
+  "info": {"title": "My API", "version": "1.0"},
+  "paths": {
+    "/api/data": {
+      "get": {
+        "parameters": [
+          {
+            "name": "X-API-Key",
+            "in": "header",
+            "schema": {"type": "string"},
+            "example": "sk-proj-abc123def456ghi789jkl"
+          }
+        ]
+      }
+    }
+  },
+  "components": {
+    "securitySchemes": {
+      "apiKey": {
+        "type": "apiKey",
+        "in": "header",
+        "name": "Authorization",
+        "default": "Bearer sk-live-xxxxxxxxxxxxxxxxxxxx"
+      }
+    }
+  }
+}`
+
+const swaggerCleanFixtureJSON = `{
+  "openapi": "3.0.0",
+  "info": {"title": "My API", "version": "1.0"},
+  "paths": {
+    "/api/data": {
+      "get": {
+        "parameters": [
+          {
+            "name": "limit",
+            "in": "query",
+            "schema": {"type": "integer"},
+            "example": 10
+          }
+        ]
+      }
+    }
+  }
+}`
+
+func TestSwagger_Sweep_ExtractsFindings(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(swaggerFixtureJSON))
+	}))
+	defer srv.Close()
+
+	src := &SwaggerSource{
+		BaseURL:  srv.URL,
+		Registry: swaggerTestRegistry(),
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 64)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	if err := src.Sweep(ctx, "", out); err != nil {
+		t.Fatalf("Sweep err: %v", err)
+	}
+	close(out)
+
+	var findings []recon.Finding
+	for f := range out {
+		findings = append(findings, f)
+	}
+	if len(findings) == 0 {
+		t.Fatal("expected at least one finding")
+	}
+	for _, f := range findings {
+		if f.SourceType != "recon:swagger" {
+			t.Errorf("unexpected SourceType: %s", f.SourceType)
+		}
+		if f.Confidence != "medium" {
+			t.Errorf("unexpected Confidence: %s", f.Confidence)
+		}
+	}
+}
+
+func TestSwagger_Sweep_NoFindings_OnCleanDoc(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(swaggerCleanFixtureJSON))
+	}))
+	defer srv.Close()
+
+	src := &SwaggerSource{
+		BaseURL:  srv.URL,
+		Registry: swaggerTestRegistry(),
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 64)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	if err := src.Sweep(ctx, "", out); err != nil {
+		t.Fatalf("Sweep err: %v", err)
+	}
+	close(out)
+
+	var count int
+	for range out {
+		count++
+	}
+	if count != 0 {
+		t.Errorf("expected 0 findings, got %d", count)
+	}
+}
+
+func TestSwagger_Sweep_CtxCancelled(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		time.Sleep(500 * time.Millisecond)
+		_, _ = w.Write([]byte(swaggerFixtureJSON))
+	}))
+	defer srv.Close()
+
+	src := &SwaggerSource{
+		BaseURL:  srv.URL,
+		Registry: swaggerTestRegistry(),
+		Limiters: recon.NewLimiterRegistry(),
+		Client:   NewClient(),
+	}
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	out := make(chan recon.Finding, 4)
+	if err := src.Sweep(ctx, "", out); err == nil {
+		t.Fatal("expected ctx error")
+	}
+}
+
+func TestSwagger_EnabledAlwaysTrue(t *testing.T) {
+	s := &SwaggerSource{}
+	if !s.Enabled(recon.Config{}) {
+		t.Fatal("expected Enabled=true")
+	}
+}
+
+func TestSwagger_NameAndRate(t *testing.T) {
+	s := &SwaggerSource{}
+	if s.Name() != "swagger" {
+		t.Errorf("unexpected name: %s", s.Name())
+	}
+	if s.Burst() != 2 {
+		t.Errorf("burst: %d", s.Burst())
+	}
+	if !s.RespectsRobots() {
+		t.Error("expected RespectsRobots=true")
+	}
+}
--- a/pkg/recon/sources/webpack.go
+++ b/pkg/recon/sources/webpack.go
@@ -0,0 +1,109 @@
+package sources
+
+import (
+	"context"
+	"fmt"
+	"io"
+	"net/http"
+	"regexp"
+	"time"
+
+	"golang.org/x/time/rate"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+// WebpackSource probes for Webpack/Vite build artifacts that contain inlined
+// environment variables. Bundlers like Webpack and Vite inline process.env.*
+// values at build time, frequently shipping API keys to production bundles.
+type WebpackSource struct {
+	BaseURL  string
+	Registry *providers.Registry
+	Limiters *recon.LimiterRegistry
+	Client   *Client
+}
+
+var _ recon.ReconSource = (*WebpackSource)(nil)
+
+func (s *WebpackSource) Name() string              { return "webpack" }
+func (s *WebpackSource) RateLimit() rate.Limit      { return rate.Every(3 * time.Second) }
+func (s *WebpackSource) Burst() int                 { return 2 }
+func (s *WebpackSource) RespectsRobots() bool       { return true }
+func (s *WebpackSource) Enabled(_ recon.Config) bool { return true }
+
+// envVarPattern matches inlined environment variable patterns from bundlers.
+var envVarPattern = regexp.MustCompile(`(?i)(NEXT_PUBLIC_|REACT_APP_|VITE_|VUE_APP_|NUXT_|GATSBY_)[A-Z_]*(API[_]?KEY|SECRET|TOKEN|PASSWORD)['":\s]*[=:,]\s*['"]([a-zA-Z0-9_\-]{8,})['"]`)
+
+// webpackBundlePaths are common locations for JS bundle artifacts.
+var webpackBundlePaths = []string{
+	"/static/js/main.js",
+	"/static/js/bundle.js",
+	"/_next/static/chunks/main.js",
+	"/assets/index.js",
+	"/dist/bundle.js",
+	"/build/static/js/main.js",
+}
+
+func (s *WebpackSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
+	base := s.BaseURL
+	if base == "" {
+		return nil
+	}
+	client := s.Client
+	if client == nil {
+		client = NewClient()
+	}
+
+	queries := BuildQueries(s.Registry, "webpack")
+	if len(queries) == 0 {
+		return nil
+	}
+
+	for _, q := range queries {
+		if err := ctx.Err(); err != nil {
+			return err
+		}
+
+		for _, path := range webpackBundlePaths {
+			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
+				}
+			}
+
+			probeURL := fmt.Sprintf("%s%s", base, path)
+			req, err := http.NewRequestWithContext(ctx, http.MethodGet, probeURL, nil)
+			if err != nil {
+				continue
+			}
+
+			resp, err := client.Do(ctx, req)
+			if err != nil {
+				continue
+			}
+
+			body, err := io.ReadAll(io.LimitReader(resp.Body, 512*1024)) // 512KB max
+			_ = resp.Body.Close()
+			if err != nil {
+				continue
+			}
+
+			if envVarPattern.Match(body) {
+				out <- recon.Finding{
+					ProviderName: q,
+					Source:       probeURL,
+					SourceType:   "recon:webpack",
+					Confidence:   "medium",
+					DetectedAt:   time.Now(),
+				}
+				break // one finding per query is sufficient
+			}
+		}
+	}
+	return nil
+}
--- a/pkg/recon/sources/webpack_test.go
+++ b/pkg/recon/sources/webpack_test.go
@@ -0,0 +1,146 @@
+package sources
+
+import (
+	"context"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+	"time"
+
+	"github.com/salvacybersec/keyhunter/pkg/providers"
+	"github.com/salvacybersec/keyhunter/pkg/recon"
+)
+
+func webpackTestRegistry() *providers.Registry {
+	return providers.NewRegistryFromProviders([]providers.Provider{
+		{Name: "openai", Keywords: []string{"sk-proj-"}},
+	})
+}
+
+const webpackFixtureJS = `
+!function(e){var t={};function n(r){if(t[r])return t[r].exports}
+var config = {
+  NEXT_PUBLIC_API_KEY: "sk-proj-abc123def456ghi789jkl",
+  REACT_APP_SECRET: "super-secret-value-12345678"
+};
+module.exports = config;
+`
+
+const webpackCleanJS = `
+!function(e){var t={};function n(r){if(t[r])return t[r].exports}
+console.log("clean bundle");
+module.exports = {};
+`
+
+func TestWebpack_Sweep_ExtractsFindings(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/javascript")
+		_, _ = w.Write([]byte(webpackFixtureJS))
+	}))
+	defer srv.Close()
+
+	src := &WebpackSource{
+		BaseURL:  srv.URL,
+		Registry: webpackTestRegistry(),
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 64)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	if err := src.Sweep(ctx, "", out); err != nil {
+		t.Fatalf("Sweep err: %v", err)
+	}
+	close(out)
+
+	var findings []recon.Finding
+	for f := range out {
+		findings = append(findings, f)
+	}
+	if len(findings) == 0 {
+		t.Fatal("expected at least one finding")
+	}
+	for _, f := range findings {
+		if f.SourceType != "recon:webpack" {
+			t.Errorf("unexpected SourceType: %s", f.SourceType)
+		}
+		if f.Confidence != "medium" {
+			t.Errorf("unexpected Confidence: %s", f.Confidence)
+		}
+	}
+}
+
+func TestWebpack_Sweep_NoFindings_OnCleanBundle(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/javascript")
+		_, _ = w.Write([]byte(webpackCleanJS))
+	}))
+	defer srv.Close()
+
+	src := &WebpackSource{
+		BaseURL:  srv.URL,
+		Registry: webpackTestRegistry(),
+		Client:   NewClient(),
+	}
+
+	out := make(chan recon.Finding, 64)
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	if err := src.Sweep(ctx, "", out); err != nil {
+		t.Fatalf("Sweep err: %v", err)
+	}
+	close(out)
+
+	var count int
+	for range out {
+		count++
+	}
+	if count != 0 {
+		t.Errorf("expected 0 findings, got %d", count)
+	}
+}
+
+func TestWebpack_Sweep_CtxCancelled(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		time.Sleep(500 * time.Millisecond)
+		_, _ = w.Write([]byte(webpackFixtureJS))
+	}))
+	defer srv.Close()
+
+	src := &WebpackSource{
+		BaseURL:  srv.URL,
+		Registry: webpackTestRegistry(),
+		Limiters: recon.NewLimiterRegistry(),
+		Client:   NewClient(),
+	}
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	out := make(chan recon.Finding, 4)
+	if err := src.Sweep(ctx, "", out); err == nil {
+		t.Fatal("expected ctx error")
+	}
+}
+
+func TestWebpack_EnabledAlwaysTrue(t *testing.T) {
+	s := &WebpackSource{}
+	if !s.Enabled(recon.Config{}) {
+		t.Fatal("expected Enabled=true")
+	}
+}
+
+func TestWebpack_NameAndRate(t *testing.T) {
+	s := &WebpackSource{}
+	if s.Name() != "webpack" {
+		t.Errorf("unexpected name: %s", s.Name())
+	}
+	if s.Burst() != 2 {
+		t.Errorf("burst: %d", s.Burst())
+	}
+	if !s.RespectsRobots() {
+		t.Error("expected RespectsRobots=true")
+	}
+}