230 lines
6.5 KiB
Go
230 lines
6.5 KiB
Go
package sources
|
|
|
|
import (
|
|
<<<<<<< HEAD
|
|
"bufio"
|
|
"context"
|
|
"fmt"
|
|
"net/http"
|
|
"net/url"
|
|
"strings"
|
|
=======
|
|
"context"
|
|
"encoding/json"
|
|
"fmt"
|
|
"io"
|
|
"net/http"
|
|
>>>>>>> worktree-agent-adad8c10
|
|
"time"
|
|
|
|
"golang.org/x/time/rate"
|
|
|
|
"github.com/salvacybersec/keyhunter/pkg/providers"
|
|
"github.com/salvacybersec/keyhunter/pkg/recon"
|
|
)
|
|
|
|
<<<<<<< HEAD
|
|
// WaybackMachineSource implements recon.ReconSource against the Wayback Machine
|
|
// CDX Server API. It queries web.archive.org/cdx/search/cdx for historical
|
|
// snapshots of pages matching provider keywords (e.g. domains known to host
|
|
// API key documentation or configuration files).
|
|
//
|
|
// RECON-ARCH-01: Each matching CDX record yields a Finding pointing at the
|
|
// archived snapshot URL. The source is credentialless and always enabled.
|
|
type WaybackMachineSource struct {
|
|
// BaseURL defaults to https://web.archive.org. 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 WaybackMachineSource satisfies recon.ReconSource.
|
|
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 }
|
|
|
|
// Enabled always returns true: CDX API is unauthenticated.
|
|
func (s *WaybackMachineSource) Enabled(_ recon.Config) bool { return true }
|
|
|
|
// Sweep iterates provider keywords, queries the CDX API for each, and emits
|
|
// a Finding for every archived snapshot URL returned. The CDX API returns
|
|
// plain-text lines with space-separated fields; we extract the original URL
|
|
// and timestamp to construct the full Wayback snapshot link.
|
|
=======
|
|
// 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 }
|
|
|
|
>>>>>>> worktree-agent-adad8c10
|
|
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
|
|
}
|
|
}
|
|
|
|
<<<<<<< HEAD
|
|
// CDX API: output=text, fl=timestamp,original limits response to two fields per line.
|
|
// limit=50 keeps the response bounded per keyword.
|
|
endpoint := fmt.Sprintf("%s/cdx/search/cdx?url=*&output=text&fl=timestamp,original&limit=50&matchType=prefix&filter=statuscode:200&query=%s",
|
|
base, url.QueryEscape(q))
|
|
req, err := http.NewRequestWithContext(ctx, http.MethodGet, endpoint, nil)
|
|
if err != nil {
|
|
return fmt.Errorf("wayback: build req: %w", err)
|
|
}
|
|
req.Header.Set("Accept", "text/plain")
|
|
|
|
resp, err := client.Do(ctx, req)
|
|
if err != nil {
|
|
// Non-fatal: skip this keyword on transient errors.
|
|
continue
|
|
}
|
|
|
|
scanner := bufio.NewScanner(resp.Body)
|
|
for scanner.Scan() {
|
|
line := strings.TrimSpace(scanner.Text())
|
|
if line == "" {
|
|
continue
|
|
}
|
|
// CDX text output: "timestamp original-url"
|
|
parts := strings.SplitN(line, " ", 2)
|
|
if len(parts) < 2 {
|
|
continue
|
|
}
|
|
ts := parts[0]
|
|
origURL := parts[1]
|
|
|
|
snapshotURL := fmt.Sprintf("%s/web/%s/%s", base, ts, origURL)
|
|
|
|
f := recon.Finding{
|
|
ProviderName: "",
|
|
Source: snapshotURL,
|
|
SourceType: "recon:wayback",
|
|
Confidence: "low",
|
|
DetectedAt: time.Now(),
|
|
}
|
|
select {
|
|
case out <- f:
|
|
case <-ctx.Done():
|
|
_ = resp.Body.Close()
|
|
return ctx.Err()
|
|
}
|
|
}
|
|
_ = resp.Body.Close()
|
|
=======
|
|
// 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(),
|
|
}
|
|
}
|
|
}
|
|
>>>>>>> worktree-agent-adad8c10
|
|
}
|
|
return nil
|
|
}
|