keyhunter/pkg/recon/sources/travisci.go

package sources

import (
	"context"
	"encoding/json"
	"fmt"
<<<<<<< HEAD
	"net/http"
	"net/url"
	"strings"
=======
	"io"
	"net/http"
	"regexp"
>>>>>>> worktree-agent-adad8c10
	"time"

	"golang.org/x/time/rate"

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

<<<<<<< HEAD
// TravisCISource searches public Travis CI build logs for leaked API keys.
// It queries the Travis CI API v3 /builds endpoint for builds matching
// provider keywords. No authentication required for public repositories.
=======
// 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.
>>>>>>> worktree-agent-adad8c10
type TravisCISource struct {
	BaseURL  string
	Registry *providers.Registry
	Limiters *recon.LimiterRegistry
	Client   *Client
}

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

<<<<<<< HEAD
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 1 }
func (s *TravisCISource) RespectsRobots() bool     { return true }
func (s *TravisCISource) Enabled(_ recon.Config) bool { return true }

=======
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.
>>>>>>> worktree-agent-adad8c10
type travisBuildResponse struct {
	Builds []travisBuild `json:"builds"`
}

type travisBuild struct {
<<<<<<< HEAD
	ID         int              `json:"id"`
	State      string           `json:"state"`
	Repository travisRepository `json:"repository"`
}

type travisRepository struct {
	Slug string `json:"slug"`
}
=======
	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,})['"]?`)
>>>>>>> worktree-agent-adad8c10

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")
<<<<<<< HEAD
	kwIndex := travisKeywordIndex(s.Registry)
=======
	if len(queries) == 0 {
		return nil
	}
>>>>>>> worktree-agent-adad8c10

	for _, q := range queries {
		if err := ctx.Err(); err != nil {
			return err
		}
<<<<<<< HEAD
=======

>>>>>>> worktree-agent-adad8c10
		if s.Limiters != nil {
			if err := s.Limiters.Wait(ctx, s.Name(), s.RateLimit(), s.Burst(), false); err != nil {
				return err
			}
		}

<<<<<<< HEAD
		endpoint := fmt.Sprintf("%s/builds?limit=20&sort_by=finished_at:desc&state=passed&event_type=push",
			base)
		req, err := http.NewRequestWithContext(ctx, http.MethodGet, endpoint, nil)
		if err != nil {
			return fmt.Errorf("travisci: build request: %w", 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
>>>>>>> worktree-agent-adad8c10
		}
		req.Header.Set("Travis-API-Version", "3")
		req.Header.Set("Accept", "application/json")

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

<<<<<<< HEAD
		var result travisBuildResponse
		decErr := json.NewDecoder(resp.Body).Decode(&result)
		_ = resp.Body.Close()
		if decErr != nil {
			continue
		}

		provName := kwIndex[strings.ToLower(q)]
		for _, build := range result.Builds {
			source := fmt.Sprintf("https://app.travis-ci.com/%s/builds/%d",
				url.PathEscape(build.Repository.Slug), build.ID)
			f := recon.Finding{
				ProviderName: provName,
				Confidence:   "low",
				Source:       source,
				SourceType:   "recon:travisci",
				DetectedAt:   time.Now(),
			}
			select {
			case out <- f:
			case <-ctx.Done():
				return ctx.Err()
=======
		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(),
				}
>>>>>>> worktree-agent-adad8c10
			}
		}
	}
	return nil
}
<<<<<<< HEAD

func travisKeywordIndex(reg *providers.Registry) map[string]string {
	m := make(map[string]string)
	if reg == nil {
		return m
	}
	for _, p := range reg.List() {
		for _, k := range p.Keywords {
			kl := strings.ToLower(strings.TrimSpace(k))
			if kl != "" {
				if _, exists := m[kl]; !exists {
					m[kl] = p.Name
				}
			}
		}
	}
	return m
}
=======
>>>>>>> worktree-agent-adad8c10