feat(go): add bounded hot-path enrichment

go-agent/internal/enrich/enrich.go

@@ -0,0 +1,166 @@
+// SPDX-License-Identifier: GPL-3.0-or-later
+
+// Package enrich builds bounded, no-I/O post-alert context from data already
+// captured for a snapshot. Filesystem hashing, package ownership, and integrity
+// collectors are intentionally excluded from this hot-path stage.
+package enrich
+
+import (
+	"regexp"
+	"sort"
+	"strings"
+
+	"codeberg.org/anassaeneroi/enodia-sentinal/go-agent/internal/model"
+	"codeberg.org/anassaeneroi/enodia-sentinal/go-agent/internal/netutil"
+)
+
+const (
+	maxAlerts     = 128
+	maxProcesses  = 64
+	maxIndicators = 128
+)
+
+var (
+	ipv4Pattern = regexp.MustCompile(`(?:[0-9]{1,3}\.){3}[0-9]{1,3}`)
+	pathPattern = regexp.MustCompile(`/[^\s,\])]+`)
+)
+
+type Process struct {
+	PID      int
+	Comm     string
+	Exe      string
+	PPID     int
+	PPIDComm string
+}
+
+// Build returns the stable Python enrichment keys without invoking any slow
+// collectors. Input and indicator caps keep attacker-controlled details from
+// causing unbounded snapshot work or output.
+func Build(alerts []model.Alert, processes []Process, lineage []int) map[string]any {
+	return map[string]any{
+		"source":        "enodia-sentinel-go",
+		"processes":     processContext(processes),
+		"remotes":       remoteContext(alerts),
+		"files":         fileCandidates(alerts),
+		"recent_writes": []string{},
+		"lineage":       lineageContext(processes, lineage),
+		"integrity":     map[string]any{},
+	}
+}
+
+func processContext(processes []Process) []map[string]any {
+	if len(processes) > maxProcesses {
+		processes = processes[:maxProcesses]
+	}
+	result := make([]map[string]any, 0, len(processes))
+	for _, process := range processes {
+		chain := []map[string]any{}
+		if process.PPID > 0 {
+			chain = append(chain, map[string]any{
+				"pid": process.PPID, "comm": process.PPIDComm, "ppid": 0,
+			})
+		}
+		result = append(result, map[string]any{
+			"pid": process.PID, "comm": process.Comm,
+			"exe": cleanPath(process.Exe), "package_owner": nil,
+			"exe_sha256": nil, "parent_chain": chain,
+		})
+	}
+	return result
+}
+
+func remoteContext(alerts []model.Alert) []map[string]any {
+	values := map[string]bool{}
+	for _, alert := range cappedAlerts(alerts) {
+		for _, value := range ipv4Pattern.FindAllString(alert.Detail, -1) {
+			if len(values) >= maxIndicators {
+				break
+			}
+			values[value] = true
+		}
+	}
+	keys := sortedKeys(values)
+	result := make([]map[string]any, 0, len(keys))
+	for _, value := range keys {
+		classification := "reserved"
+		address, ok := netutil.ParseAddr(value)
+		if !ok {
+			classification = "invalid"
+		} else if netutil.IsPublicIP(value) {
+			classification = "public"
+		} else if address.IsLoopback() {
+			classification = "loopback"
+		} else if address.IsPrivate() {
+			classification = "private"
+		} else if address.IsLinkLocalUnicast() {
+			classification = "link-local"
+		}
+		result = append(result, map[string]any{"ip": value, "classification": classification})
+	}
+	return result
+}
+
+func fileCandidates(alerts []model.Alert) []map[string]any {
+	values := map[string]bool{}
+	for _, alert := range cappedAlerts(alerts) {
+		for _, text := range []string{alert.Detail, alert.Key} {
+			for _, value := range pathPattern.FindAllString(text, -1) {
+				if len(values) >= maxIndicators {
+					break
+				}
+				values[cleanPath(strings.TrimSuffix(value, ":"))] = true
+			}
+		}
+	}
+	keys := sortedKeys(values)
+	result := make([]map[string]any, 0, len(keys))
+	for _, value := range keys {
+		// Existence/ownership is unknown until the deferred collector runs.
+		result = append(result, map[string]any{
+			"path": value, "exists": nil, "mode": nil, "uid": nil,
+			"gid": nil, "package_owner": nil,
+		})
+	}
+	return result
+}
+
+func lineageContext(processes []Process, lineage []int) []map[string]any {
+	known := map[int]string{}
+	for _, process := range processes {
+		known[process.PID] = process.Comm
+		if process.PPID > 0 && known[process.PPID] == "" {
+			known[process.PPID] = process.PPIDComm
+		}
+	}
+	values := append([]int(nil), lineage...)
+	sort.Ints(values)
+	if len(values) > maxProcesses*4 {
+		values = values[:maxProcesses*4]
+	}
+	result := make([]map[string]any, 0, len(values))
+	for _, pid := range values {
+		comm, present := known[pid]
+		result = append(result, map[string]any{"pid": pid, "comm": comm, "present": present})
+	}
+	return result
+}
+
+func cappedAlerts(alerts []model.Alert) []model.Alert {
+	if len(alerts) > maxAlerts {
+		return alerts[:maxAlerts]
+	}
+	return alerts
+}
+
+func sortedKeys(values map[string]bool) []string {
+	result := make([]string, 0, len(values))
+	for value := range values {
+		if value != "" {
+			result = append(result, value)
+		}
+	}
+	sort.Strings(result)
+	return result
+}
+
+func cleanPath(value string) string { return strings.ReplaceAll(value, " (deleted)", "") }

go-agent/internal/enrich/enrich_test.go

@@ -0,0 +1,56 @@
+// SPDX-License-Identifier: GPL-3.0-or-later
+
+package enrich
+
+import (
+	"testing"
+
+	"codeberg.org/anassaeneroi/enodia-sentinal/go-agent/internal/model"
+)
+
+func TestBuildAddsBoundedProcessRemotePathAndLineageContext(t *testing.T) {
+	result := Build([]model.Alert{{
+		SID: 100010, Signature: "reverse_shell", Key: "persist:/etc/cron.d/dropper",
+		Detail: "pid=42 peer=[8.8.8.8:4444] wrote /tmp/dropper (deleted)",
+	}}, []Process{{PID: 42, Comm: "bash", Exe: "/tmp/dropper (deleted)", PPID: 10, PPIDComm: "nginx"}}, []int{42, 10})
+	processes := result["processes"].([]map[string]any)
+	if len(processes) != 1 || processes[0]["exe"] != "/tmp/dropper" ||
+		processes[0]["package_owner"] != nil || processes[0]["exe_sha256"] != nil {
+		t.Fatalf("processes=%#v", processes)
+	}
+	chain := processes[0]["parent_chain"].([]map[string]any)
+	if len(chain) != 1 || chain[0]["pid"] != 10 || chain[0]["comm"] != "nginx" {
+		t.Fatalf("chain=%#v", chain)
+	}
+	remotes := result["remotes"].([]map[string]any)
+	if len(remotes) != 1 || remotes[0]["ip"] != "8.8.8.8" || remotes[0]["classification"] != "public" {
+		t.Fatalf("remotes=%#v", remotes)
+	}
+	files := result["files"].([]map[string]any)
+	if len(files) != 2 || files[0]["path"] != "/etc/cron.d/dropper" || files[1]["path"] != "/tmp/dropper" {
+		t.Fatalf("files=%#v", files)
+	}
+	lineage := result["lineage"].([]map[string]any)
+	if len(lineage) != 2 || lineage[0]["pid"] != 10 || lineage[0]["comm"] != "nginx" {
+		t.Fatalf("lineage=%#v", lineage)
+	}
+}
+
+func TestBuildClassifiesPrivateAndInvalidAddresses(t *testing.T) {
+	result := Build([]model.Alert{{Detail: "peers 10.1.2.3 and 999.1.2.3"}}, nil, nil)
+	remotes := result["remotes"].([]map[string]any)
+	if len(remotes) != 2 || remotes[0]["classification"] != "private" || remotes[1]["classification"] != "invalid" {
+		t.Fatalf("remotes=%#v", remotes)
+	}
+}
+
+func TestBuildCapsAttackerControlledIndicators(t *testing.T) {
+	alerts := make([]model.Alert, 300)
+	for index := range alerts {
+		alerts[index].Detail = "/tmp/path " + "1.2.3.4"
+	}
+	result := Build(alerts, nil, nil)
+	if len(result["remotes"].([]map[string]any)) > maxIndicators || len(result["files"].([]map[string]any)) > maxIndicators {
+		t.Fatal("indicator caps were not applied")
+	}
+}

go-agent/internal/snapshot/store.go

@@ -17,6 +17,7 @@ import (
 	"time"
 
 	"codeberg.org/anassaeneroi/enodia-sentinal/go-agent/internal/correlation"
+	"codeberg.org/anassaeneroi/enodia-sentinal/go-agent/internal/enrich"
 	"codeberg.org/anassaeneroi/enodia-sentinal/go-agent/internal/incident"
 	"codeberg.org/anassaeneroi/enodia-sentinal/go-agent/internal/model"
 )
@@ -54,14 +55,15 @@ type Stats struct {
 }
 
 type Store struct {
-	Dir        string
-	ProcRoot   string
-	MaxCount   int
-	MaxAgeDays int
-	Now        func() time.Time
-	Collect    func(int) ProcessDetail
-	ParentOf   func(int) int
-	Incidents  *incident.Store
+	Dir          string
+	ProcRoot     string
+	MaxCount     int
+	MaxAgeDays   int
+	Now          func() time.Time
+	Collect      func(int) ProcessDetail
+	ParentOf     func(int) int
+	Incidents    *incident.Store
+	LineageDepth int
 
 	mu sync.Mutex
 }
@@ -97,7 +99,7 @@ func New(dir, procRoot string, maxCount, maxAgeDays int) (*Store, error) {
 	if err := os.Remove(probePath); err != nil {
 		return nil, fmt.Errorf("remove snapshot preflight: %w", err)
 	}
-	store := &Store{Dir: dir, ProcRoot: procRoot, MaxCount: maxCount, MaxAgeDays: maxAgeDays, Now: time.Now}
+	store := &Store{Dir: dir, ProcRoot: procRoot, MaxCount: maxCount, MaxAgeDays: maxAgeDays, Now: time.Now, LineageDepth: 8}
 	store.Collect = func(pid int) ProcessDetail { return collectProcess(procRoot, pid) }
 	store.ParentOf = func(pid int) int { return readParent(procRoot, pid) }
 	if err := store.Prune(); err != nil {
@@ -114,6 +116,7 @@ func (s *Store) ConfigureIncidents(enabled bool, window, lineageDepth int) error
 		return err
 	}
 	s.Incidents = tracker
+	s.LineageDepth = lineageDepth
 	return nil
 }
 
@@ -160,9 +163,9 @@ func (s *Store) CaptureEvent(event map[string]any) (string, error) {
 	report.Alerts = append(report.Alerts, alert)
 	report.Severity = maxSeverity(report.Severity, alert.Severity)
 	report.Processes = s.collectProcesses(report.Alerts)
+	pids := alertPIDs(report.Alerts)
+	lineage := correlation.LineageOf(pids, s.ParentOf, s.LineageDepth)
 	if s.Incidents != nil {
-		pids := alertPIDs(report.Alerts)
-		lineage := correlation.LineageOf(pids, s.ParentOf, s.Incidents.Depth)
 		preferred := ""
 		if report.IncidentID != nil {
 			preferred = *report.IncidentID
@@ -173,9 +176,7 @@ func (s *Store) CaptureEvent(event map[string]any) (string, error) {
 			report.IncidentID = &id
 		}
 	}
-	if report.Enrichment == nil {
-		report.Enrichment = map[string]any{"source": "enodia-sentinel-go"}
-	}
+	report.Enrichment = enrich.Build(report.Alerts, enrichmentProcesses(report.Processes), lineageValues(lineage))
 	if err := writeJSONAtomic(jsonPath, report); err != nil {
 		return "", err
 	}
@@ -212,6 +213,28 @@ func alertPIDs(alerts []model.Alert) []int {
 	return pids
 }
 
+func enrichmentProcesses(processes []ProcessDetail) []enrich.Process {
+	result := make([]enrich.Process, 0, len(processes))
+	for _, process := range processes {
+		result = append(result, enrich.Process{
+			PID: process.PID, Comm: process.Comm, Exe: process.Exe,
+			PPID: process.PPID, PPIDComm: process.PPIDComm,
+		})
+	}
+	return result
+}
+
+func lineageValues(lineage map[int]bool) []int {
+	result := make([]int, 0, len(lineage))
+	for pid, included := range lineage {
+		if included {
+			result = append(result, pid)
+		}
+	}
+	sort.Ints(result)
+	return result
+}
+
 // SnapshotStats prunes first, then derives status claims from retained,
 // parseable snapshot files. Calling it for every status emission keeps the age
 // bound active even when a quiet host produces no new alert snapshots.