keyhunter/pkg/recon/sources/travisci.go

package sources

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

	"golang.org/x/time/rate"

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

// TravisCISource scrapes public Travis CI build logs for leaked API keys.
// Travis CI exposes build logs publicly by default for open-source projects.
// Developers frequently print environment variables or use secrets insecurely
// in CI scripts, causing API keys to appear in build output.
type TravisCISource struct {
	BaseURL  string
	Registry *providers.Registry
	Limiters *recon.LimiterRegistry
	Client   *Client
}

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

func (s *TravisCISource) Name() string              { return "travisci" }
func (s *TravisCISource) RateLimit() rate.Limit      { return rate.Every(3 * time.Second) }
func (s *TravisCISource) Burst() int                 { return 2 }
func (s *TravisCISource) RespectsRobots() bool       { return false }
func (s *TravisCISource) Enabled(_ recon.Config) bool { return true }

// travisBuildResponse represents the Travis CI API builds response.
type travisBuildResponse struct {
	Builds []travisBuild `json:"builds"`
}

type travisBuild struct {
	ID    int    `json:"id"`
	State string `json:"state"`
}

// ciLogKeyPattern matches API key patterns commonly leaked in CI logs.
var ciLogKeyPattern = regexp.MustCompile(`(?i)(api[_-]?key|secret[_-]?key|token|password|credential|auth[_-]?token)['":\s]*[=:]\s*['"]?([a-zA-Z0-9_\-]{16,})['"]?`)

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

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

		// Search for builds related to the query keyword.
		searchURL := fmt.Sprintf("%s/builds?search=%s&limit=5", base, q)
		req, err := http.NewRequestWithContext(ctx, http.MethodGet, searchURL, nil)
		if err != nil {
			continue
		}
		req.Header.Set("Travis-API-Version", "3")
		req.Header.Set("Accept", "application/json")

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

		var builds travisBuildResponse
		if err := json.NewDecoder(resp.Body).Decode(&builds); err != nil {
			_ = resp.Body.Close()
			continue
		}
		_ = resp.Body.Close()

		for _, b := range builds.Builds {
			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
				}
			}

			// Fetch the build log.
			logURL := fmt.Sprintf("%s/builds/%d/log", base, b.ID)
			logReq, err := http.NewRequestWithContext(ctx, http.MethodGet, logURL, nil)
			if err != nil {
				continue
			}
			logReq.Header.Set("Travis-API-Version", "3")
			logReq.Header.Set("Accept", "text/plain")

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

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

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