update

backend/pkg/agentcli/guardrail/sanitize.go

@@ -0,0 +1,108 @@
+// Package guardrail 提供 Agent CLI 输入输出治理能力。 // Logan.Li
+package guardrail
+
+import (
+	"crypto/rand"
+	"encoding/hex"
+	"encoding/json"
+	"regexp"
+	"strings"
+)
+
+var sensitivePatterns = []*regexp.Regexp{
+	regexp.MustCompile(`(?i)(bearer\s+)[a-z0-9._~+/=-]{16,}`),
+	regexp.MustCompile(`(?i)((?:password|passwd|pwd|secret|token|api[_-]?key|access[_-]?key|secret[_-]?key)\s*[:=]\s*)[^\s'"&]+`),
+	regexp.MustCompile(`(?i)((?:mysql|postgres|postgresql|redis|mongodb)://[^:\s/@]+:)[^@\s]+(@)`),
+	regexp.MustCompile(`(?is)-----BEGIN [A-Z ]*PRIVATE KEY-----.*?-----END [A-Z ]*PRIVATE KEY-----`),
+	regexp.MustCompile(`(?i)((?:Cookie|Set-Cookie):\s*)[^\r\n]+`),
+	regexp.MustCompile(`(?i)(AKIA[0-9A-Z]{16})`),
+}
+
+var sensitiveKeys = map[string]bool{
+	"password": true, "passwd": true, "pwd": true, "secret": true,
+	"token": true, "api_key": true, "apikey": true, "access_key": true,
+	"secret_key": true, "private_key": true, "cookie": true,
+}
+
+// SanitizeText 脱敏工具输出、LLM 文本和错误信息，避免敏感信息进入前端/DB/记忆。 // Logan.Li
+func SanitizeText(input string) string {
+	if input == "" {
+		return ""
+	}
+	out := input
+	for _, re := range sensitivePatterns {
+		out = re.ReplaceAllStringFunc(out, func(match string) string {
+			if strings.Contains(match, "PRIVATE KEY") {
+				return "[REDACTED_PRIVATE_KEY]"
+			}
+			sub := re.FindStringSubmatch(match)
+			if len(sub) >= 3 {
+				return sub[1] + "[REDACTED]" + sub[2]
+			}
+			if len(sub) >= 2 {
+				return sub[1] + "[REDACTED]"
+			}
+			return "[REDACTED]"
+		})
+	}
+	return out
+}
+
+// SanitizeJSON 脱敏 JSON 文本；非法 JSON 时退回文本脱敏。 // Logan.Li
+func SanitizeJSON(raw string) string {
+	if strings.TrimSpace(raw) == "" {
+		return raw
+	}
+	var v interface{}
+	if err := json.Unmarshal([]byte(raw), &v); err != nil {
+		return SanitizeText(raw)
+	}
+	v = sanitizeValue(v)
+	b, err := json.Marshal(v)
+	if err != nil {
+		return SanitizeText(raw)
+	}
+	return string(b)
+}
+
+func sanitizeValue(v interface{}) interface{} {
+	switch val := v.(type) {
+	case map[string]interface{}:
+		out := make(map[string]interface{}, len(val))
+		for k, item := range val {
+			if sensitiveKeys[strings.ToLower(strings.ReplaceAll(k, "-", "_"))] {
+				out[k] = "[REDACTED]"
+				continue
+			}
+			out[k] = sanitizeValue(item)
+		}
+		return out
+	case []interface{}:
+		for i := range val {
+			val[i] = sanitizeValue(val[i])
+		}
+		return val
+	case string:
+		return SanitizeText(val)
+	default:
+		return val
+	}
+}
+
+// NewTraceID 生成短 trace id，便于串联一次 Agent 任务。 // Logan.Li
+func NewTraceID() string {
+	return "tr-" + randomHex(12)
+}
+
+// NewSpanID 生成短 span id，便于串联工具和消息事件。 // Logan.Li
+func NewSpanID() string {
+	return "sp-" + randomHex(8)
+}
+
+func randomHex(n int) string {
+	buf := make([]byte, n)
+	if _, err := rand.Read(buf); err != nil {
+		return "0000000000000000"
+	}
+	return hex.EncodeToString(buf)
+}

backend/pkg/agentcli/guardrail/sanitize_test.go

@@ -0,0 +1,31 @@
+package guardrail
+
+import (
+	"strings"
+	"testing"
+)
+
+func TestSanitizeTextMasksCommonSecrets(t *testing.T) {
+	input := "Authorization: Bearer abcdefghijklmnopqrstuvwxyz password=secret123 mysql://root:pass@db/prod"
+	got := SanitizeText(input)
+	for _, leak := range []string{"abcdefghijklmnopqrstuvwxyz", "secret123", "root:pass@db"} {
+		if strings.Contains(got, leak) {
+			t.Fatalf("敏感信息未脱敏: %s in %s", leak, got)
+		}
+	}
+	if !strings.Contains(got, "[REDACTED]") {
+		t.Fatalf("脱敏标记缺失: %s", got)
+	}
+}
+
+func TestSanitizeJSONMasksSensitiveKeys(t *testing.T) {
+	got := SanitizeJSON(`{"command":"echo ok","password":"p1","nested":{"api_key":"k1"}}`)
+	for _, leak := range []string{"p1", "k1"} {
+		if strings.Contains(got, leak) {
+			t.Fatalf("JSON 敏感字段未脱敏: %s in %s", leak, got)
+		}
+	}
+	if !strings.Contains(got, `"command":"echo ok"`) {
+		t.Fatalf("非敏感字段不应被改写: %s", got)
+	}
+}