keyhunter/pkg/recon/sources/githubactions.go

package sources

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

	"golang.org/x/time/rate"

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

// GitHubActionsSource searches GitHub Actions workflow run logs for leaked API
<<<<<<< HEAD
// keys. It queries the GitHub REST API for workflow runs matching provider
// keywords (via the repository search endpoint) and emits findings for each
// matching run. Requires a GitHub token (same as GitHubSource).
=======
// keys. Workflow logs are public for public repositories and frequently contain
// accidentally printed secrets, debug output with credentials, or insecure
// echo statements that expose environment variables.
>>>>>>> worktree-agent-adad8c10
type GitHubActionsSource struct {
	Token    string
	BaseURL  string
	Registry *providers.Registry
	Limiters *recon.LimiterRegistry
<<<<<<< HEAD
	client   *Client
=======
	Client   *Client
>>>>>>> worktree-agent-adad8c10
}

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

<<<<<<< HEAD
func (s *GitHubActionsSource) Name() string          { return "github_actions" }
func (s *GitHubActionsSource) RateLimit() rate.Limit { return rate.Every(2 * time.Second) }
func (s *GitHubActionsSource) Burst() int            { return 2 }
func (s *GitHubActionsSource) RespectsRobots() bool  { return false }
func (s *GitHubActionsSource) Enabled(_ recon.Config) bool { return s.Token != "" }

// ghActionsSearchResponse models the GitHub code search response when looking
// for workflow files containing provider keywords.
type ghActionsSearchResponse struct {
	Items []ghActionsItem `json:"items"`
}

type ghActionsItem struct {
	HTMLURL    string              `json:"html_url"`
	Repository ghActionsRepository `json:"repository"`
}

type ghActionsRepository struct {
	FullName string `json:"full_name"`
}

func (s *GitHubActionsSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
	if s.Token == "" {
		return nil
	}
=======
func (s *GitHubActionsSource) Name() string         { return "ghactions" }
func (s *GitHubActionsSource) RateLimit() rate.Limit { return rate.Every(2 * time.Second) }
func (s *GitHubActionsSource) Burst() int            { return 3 }
func (s *GitHubActionsSource) RespectsRobots() bool  { return false }

// Enabled requires a GitHub token (reuses GitHubToken from SourcesConfig).
func (s *GitHubActionsSource) Enabled(_ recon.Config) bool { return s.Token != "" }

// ghActionsRunsResponse represents the GitHub Actions workflow runs list.
type ghActionsRunsResponse struct {
	WorkflowRuns []ghActionsRun `json:"workflow_runs"`
}

type ghActionsRun struct {
	ID         int64  `json:"id"`
	LogsURL    string `json:"logs_url"`
	HTMLURL    string `json:"html_url"`
	Status     string `json:"status"`
	Conclusion string `json:"conclusion"`
}

func (s *GitHubActionsSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
>>>>>>> worktree-agent-adad8c10
	base := s.BaseURL
	if base == "" {
		base = "https://api.github.com"
	}
<<<<<<< HEAD
	if s.client == nil {
		s.client = NewClient()
	}

	queries := BuildQueries(s.Registry, "github_actions")
	kwIndex := ghActionsKeywordIndex(s.Registry)
=======
	client := s.Client
	if client == nil {
		client = NewClient()
	}

	queries := BuildQueries(s.Registry, "ghactions")
	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
		// Search for workflow YAML files referencing the keyword.
		endpoint := fmt.Sprintf("%s/search/code?q=%s+path:.github/workflows+extension:yml&per_page=20",
			base, url.QueryEscape(q))
		req, err := http.NewRequestWithContext(ctx, http.MethodGet, endpoint, nil)
		if err != nil {
			return fmt.Errorf("github_actions: build request: %w", err)
		}
		req.Header.Set("Authorization", "Bearer "+s.Token)
		req.Header.Set("Accept", "application/vnd.github+json")

		resp, err := s.client.Do(ctx, req)
		if err != nil {
			if strings.Contains(err.Error(), "unauthorized") {
				return err
			}
			continue
		}

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

		provName := kwIndex[strings.ToLower(q)]
		for _, item := range result.Items {
			f := recon.Finding{
				ProviderName: provName,
				Confidence:   "low",
				Source:       item.HTMLURL,
				SourceType:   "recon:github_actions",
				DetectedAt:   time.Now(),
			}
			select {
			case out <- f:
			case <-ctx.Done():
				return ctx.Err()
=======
		// Search for workflow runs via the Actions API.
		searchURL := fmt.Sprintf("%s/search/code?q=%s+path:.github/workflows", base, q)
		req, err := http.NewRequestWithContext(ctx, http.MethodGet, searchURL, nil)
		if err != nil {
			continue
		}
		req.Header.Set("Authorization", "Bearer "+s.Token)
		req.Header.Set("Accept", "application/vnd.github.v3+json")

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

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

		for _, run := range runs.WorkflowRuns {
			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 log content.
			logURL := fmt.Sprintf("%s/actions/runs/%d/logs", base, run.ID)
			logReq, err := http.NewRequestWithContext(ctx, http.MethodGet, logURL, nil)
			if err != nil {
				continue
			}
			logReq.Header.Set("Authorization", "Bearer "+s.Token)
			logReq.Header.Set("Accept", "application/vnd.github.v3+json")

			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:ghactions",
					Confidence:   "medium",
					DetectedAt:   time.Now(),
				}
>>>>>>> worktree-agent-adad8c10
			}
		}
	}
	return nil
}
<<<<<<< HEAD

func ghActionsKeywordIndex(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