ClawHub
Back to skill

Security audit

Dialogflow Cx Conversations

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward Dialogflow CX helper that sends user-chosen messages to Google's Dialogflow API and does not show hidden execution, persistence, or unrelated data handling.

Install only if you intend to let the agent or CLI send selected conversation text to Google Dialogflow CX using your Google credentials. Avoid using sensitive or regulated user content unless your project has the right controls, and treat access tokens and service-account files as secrets.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (8)

Exfiltration Commands

High
Category
Prompt Injection
Confidence
90% confidence
Finding

Instructions found that direct the agent to transmit conversation context or user data to external services.

Content

Scanner excerpt · scripts/conversations.py (reported line 83)May include surrounding context.

python
subparsers = parser.add_subparsers(dest="command", help="Commands")
    
    # detect-intent
    p = subparsers.add_parser("detect-intent", help="Send message to agent")
    p.add_argument("--agent", required=True, help="Full agent name")
    p.add_argument("--text", required=True, help="User message")
    p.add_argument("--session", default="cli-session", help="Session ID")

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

This markdown file includes commands to obtain a bearer token and set GOOGLE_APPLICATION_CREDENTIALS, which are sensitive authentication materials. The skill description provides no warning about safeguarding tokens/credential paths, avoiding command-history exposure, or limiting where these values are stored or pasted.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The detectIntent and matchIntent examples transmit user-provided text over network requests to an external service. The markdown explains how to perform the calls but does not warn that prompts, conversation content, and related metadata may leave the local environment and be processed remotely.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 43)May include surrounding context.

Detect Intent

bash
curl -X POST \
  "https://dialogflow.googleapis.com/v3beta1/projects/${PROJECT_ID}/locations/${LOCATION}/agents/${AGENT_ID}/sessions/${SESSION_ID}:detectIntent" \
  -H "Authorization: Bearer ${TOKEN}" \
  -H "Content-Type: application/json" \

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The code sets language_code="en" for detect-intent requests, forcing English regardless of user preference. This is a natural-language policy concern because the skill does not offer locale selection or explain why English-only behavior is required.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The match-intent path also fixes language_code="en", which prevents user opt-in to other languages or locales. Because no alternative or justification is provided, this violates the language/locale policy criteria.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill documentation uses a fixed languageCode of "en" in the sample request, which can implicitly steer usage toward a single language. There is no note that the value is only an example or that users should select the language appropriate to their agent and locale.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The matchIntent example also fixes languageCode to "en", again implying a default English-only workflow. Without an opt-in or explanatory note, this conflicts with the requirement to avoid forcing a specific language or locale.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.