keyhunter/pkg/recon/sources/wayback.go

package sources

import (
	"context"
	"encoding/json"
	"fmt"
	"io"
	"net/http"
	"time"

	"golang.org/x/time/rate"

	"github.com/salvacybersec/keyhunter/pkg/providers"
	"github.com/salvacybersec/keyhunter/pkg/recon"
)

// WaybackMachineSource searches the Internet Archive's Wayback Machine CDX API
// for archived pages that may contain leaked API keys. Developers sometimes
// remove secrets from live pages but cached versions persist in web archives.
type WaybackMachineSource struct {
	BaseURL  string
	Registry *providers.Registry
	Limiters *recon.LimiterRegistry
	Client   *Client
}

var _ recon.ReconSource = (*WaybackMachineSource)(nil)

func (s *WaybackMachineSource) Name() string              { return "wayback" }
func (s *WaybackMachineSource) RateLimit() rate.Limit      { return rate.Every(5 * time.Second) }
func (s *WaybackMachineSource) Burst() int                 { return 1 }
func (s *WaybackMachineSource) RespectsRobots() bool       { return true }
func (s *WaybackMachineSource) Enabled(_ recon.Config) bool { return true }

func (s *WaybackMachineSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
	base := s.BaseURL
	if base == "" {
		base = "https://web.archive.org"
	}
	client := s.Client
	if client == nil {
		client = NewClient()
	}

	queries := BuildQueries(s.Registry, "wayback")
	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
			}
		}

		// CDX API: search for archived URLs matching the query.
		// Filter for .env, config, and JS files that commonly contain keys.
		cdxURL := fmt.Sprintf("%s/cdx/search/cdx?url=*%s*&output=json&limit=10&fl=url,timestamp,statuscode&filter=statuscode:200", base, q)
		req, err := http.NewRequestWithContext(ctx, http.MethodGet, cdxURL, nil)
		if err != nil {
			continue
		}
		req.Header.Set("Accept", "application/json")

		resp, err := client.Do(ctx, req)
		if err != nil {
			continue
		}

		var rows [][]string
		if err := json.NewDecoder(resp.Body).Decode(&rows); err != nil {
			_ = resp.Body.Close()
			continue
		}
		_ = resp.Body.Close()

		// Skip the header row if present.
		start := 0
		if len(rows) > 0 && len(rows[0]) > 0 && rows[0][0] == "url" {
			start = 1
		}

		for _, row := range rows[start:] {
			if err := ctx.Err(); err != nil {
				return err
			}
			if len(row) < 2 {
				continue
			}
			archivedURL := row[0]
			timestamp := row[1]

			if s.Limiters != nil {
				if err := s.Limiters.Wait(ctx, s.Name(), s.RateLimit(), s.Burst(), false); err != nil {
					return err
				}
			}

			// Fetch the archived page content.
			snapshotURL := fmt.Sprintf("%s/web/%sid_/%s", base, timestamp, archivedURL)
			snapReq, err := http.NewRequestWithContext(ctx, http.MethodGet, snapshotURL, nil)
			if err != nil {
				continue
			}

			snapResp, err := client.Do(ctx, snapReq)
			if err != nil {
				continue
			}

			body, err := io.ReadAll(io.LimitReader(snapResp.Body, 256*1024))
			_ = snapResp.Body.Close()
			if err != nil {
				continue
			}

			if apiKeyPattern.Match(body) {
				out <- recon.Finding{
					ProviderName: q,
					Source:       snapshotURL,
					SourceType:   "recon:wayback",
					Confidence:   "medium",
					DetectedAt:   time.Now(),
				}
			}
		}
	}
	return nil
}