ClawHub
Back to skill

Security audit

voice-call-memory

Security checks for vulnerabilities and agentic risk

Overview

This skill does what it says, but it stores identifiable voice-call details in an external memory service without enough privacy controls.

Review before installing. Use only where you have authority to record and store caller information, provide required notice or consent, minimize or redact phone numbers and sensitive details, verify the BlueColumn/Supabase endpoint and API-key handling, and define retention, deletion, and access controls before using it with real calls.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 (2)

Missing User Warnings

High
Confidence
97% confidence
Finding
The skill explicitly instructs storing and transmitting detailed caller data, including phone numbers, call content, emotional cues, and follow-up commitments, but provides no consent, notice, minimization, retention, or access-control guidance. In a voice-call context this is especially sensitive because it creates searchable records of identifiable conversations that may violate privacy laws, internal policy, or customer expectations if collected without informed consent.

External Transmission

Medium
Category
Data Exfiltration
Content
After each call, store a structured summary with the caller, purpose, and outcome.

```bash
curl -X POST https://xkjkwqbfvkswwdmbtndo.supabase.co/functions/v1/agent-remember \
  -H "Authorization: Bearer $BLUECOLUMN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text": "VOICE CALL +12065550123 (7/31 14:20): asked about Builder plan pricing; concern about overage fees; agreed to email pricing sheet today; tone friendly but rushed; follow up Friday.", "title": "voice - +12065550123 7/31"}'
Confidence
92% confidence
Finding
The example sends sensitive caller information to an external endpoint using an API key, including a phone number, timestamp, pricing discussion, sentiment, and planned follow-up. External transmission is expected for this skill's function, but the documented payload contains unnecessary personal and behavioral detail and lacks safeguards such as redaction, consent, or environment/security requirements, making data leakage or noncompliant processing more likely.

Static analysis

No suspicious patterns detected.