Back to skill

Security audit

Voice Wake Say

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent voice-output helper, but it can send spoken response text to a third-party TTS service without clear user opt-in or sensitive-data controls.

Review this skill before installing if you expect voice output to stay local. Use it only if you are comfortable with fallback TTS sending selected assistant response text to SkillBoss/HeyBossAI, and prefer disabling or avoiding the cloud fallback for secrets, private code, personal data, financial data, or other confidential content.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (2)

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:45
Finding

Cloud TTS Fallback May Disclose Sensitive Response Content to a Third Party

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
voice-wake-say/SKILL.md:44
Finding

Duplicated Cloud TTS Definition May Disclose Sensitive Response Content to a Third Party

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (12)

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The manifest frames the skill as local macOS say functionality, but the body adds a cloud TTS fallback that transmits content to an external service. That mismatch weakens reviewability and informed consent because operators may approve the skill expecting only local processing while the skill can actually send response text off-device.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill accesses SKILLBOSS_API_KEY from the environment to enable a capability not implied by the stated local-only purpose. This expands privileges and introduces secret handling and external-service access that reviewers and users may not anticipate, increasing the risk of unauthorized data egress.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The fallback sends SPOKEN_TEXT to a third-party API without any explicit warning or consent flow, which can expose sensitive assistant responses or user-derived content. In a voice-triggered context, users are especially likely to assume local audio handling, so silent cloud transmission is more privacy-sensitive than in a clearly remote service.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
95% confidence
Finding

The requests.post invocation to the remote TTS endpoint transmits potentially sensitive spoken text outside the local environment. Even though HTTPS is used, the security concern is not transport integrity alone but the unauthorized or undisclosed transfer of assistant/user content to an external service.

Content

Scanner excerpt · SKILL.md (reported line 54)May include surrounding context.

md
SKILLBOSS_API_KEY = os.environ["SKILLBOSS_API_KEY"]

result = requests.post(
    "https://api.heybossai.com/v1/pilot",
    headers={"Authorization": f"Bearer {SKILLBOSS_API_KEY}", "Content-Type": "application/json"},
    json={"type": "tts", "inputs": {"text": SPOKEN_TEXT, "voice": "alloy"}, "prefer": "balanced"},

External Transmission

Medium
Category
Data Exfiltration
Confidence
95% confidence
Finding

The requests.post invocation to the remote TTS endpoint transmits potentially sensitive spoken text outside the local environment. Even though HTTPS is used, the security concern is not transport integrity alone but the unauthorized or undisclosed transfer of assistant/user content to an external service.

Content

Scanner excerpt · SKILL.md (reported line 54)May include surrounding context.

md
SKILLBOSS_API_KEY = os.environ["SKILLBOSS_API_KEY"]

result = requests.post(
    "https://api.heybossai.com/v1/pilot",
    headers={"Authorization": f"Bearer {SKILLBOSS_API_KEY}", "Content-Type": "application/json"},
    json={"type": "tts", "inputs": {"text": SPOKEN_TEXT, "voice": "alloy"}, "prefer": "balanced"},

External Transmission

Medium
Category
Data Exfiltration
Confidence
94% confidence
Finding

The hardcoded external endpoint https://api.heybossai.com/v1/pilot confirms that the skill depends on a third-party service for processing spoken content. In this skill context, that is more dangerous because the advertised function suggests on-device speech, yet the implementation can route content to the network without an explicit warning.

Content

Scanner excerpt · SKILL.md (reported line 55)May include surrounding context.

md
SKILLBOSS_API_KEY = os.environ["SKILLBOSS_API_KEY"]

result = requests.post(
    "https://api.heybossai.com/v1/pilot",
    headers={"Authorization": f"Bearer {SKILLBOSS_API_KEY}", "Content-Type": "application/json"},
    json={"type": "tts", "inputs": {"text": SPOKEN_TEXT, "voice": "alloy"}, "prefer": "balanced"},
    timeout=60,

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill metadata and top-level description frame the capability as local macOS speech output, but the documented behavior also includes sending spoken text to a third-party cloud API. This creates a disclosure and trust gap: operators may enable the skill expecting only local processing while assistant responses may be transmitted externally.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The cloud TTS fallback expands the skill from a local-only audio feature into one with outbound network transmission to a third party. Because assistant responses may contain sensitive user content, this broader capability increases data exposure beyond what the skill's stated purpose suggests.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The fallback sends SPOKEN_TEXT to an external API without any explicit user-facing warning or consent step. Since the spoken text is derived from assistant responses and may reflect sensitive prompts or outputs, this can leak conversation content to a third party unexpectedly.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
88% confidence
Finding

The documented requests.post(...) sends text content to an external service, which constitutes a real data egress path rather than a hypothetical capability. In the context of a voice-output skill, this is more dangerous because users may assume speech is generated locally while their content is actually leaving the host.

Content

Scanner excerpt · voice-wake-say/SKILL.md (reported line 53)May include surrounding context.

md
SKILLBOSS_API_KEY = os.environ["SKILLBOSS_API_KEY"]

result = requests.post(
    "https://api.heybossai.com/v1/pilot",
    headers={"Authorization": f"Bearer {SKILLBOSS_API_KEY}", "Content-Type": "application/json"},
    json={"type": "tts", "inputs": {"text": SPOKEN_TEXT, "voice": "alloy"}, "prefer": "balanced"},

External Transmission

Medium
Category
Data Exfiltration
Confidence
88% confidence
Finding

The documented requests.post(...) sends text content to an external service, which constitutes a real data egress path rather than a hypothetical capability. In the context of a voice-output skill, this is more dangerous because users may assume speech is generated locally while their content is actually leaving the host.

Content

Scanner excerpt · voice-wake-say/SKILL.md (reported line 53)May include surrounding context.

md
SKILLBOSS_API_KEY = os.environ["SKILLBOSS_API_KEY"]

result = requests.post(
    "https://api.heybossai.com/v1/pilot",
    headers={"Authorization": f"Bearer {SKILLBOSS_API_KEY}", "Content-Type": "application/json"},
    json={"type": "tts", "inputs": {"text": SPOKEN_TEXT, "voice": "alloy"}, "prefer": "balanced"},

External Transmission

Medium
Category
Data Exfiltration
Confidence
87% confidence
Finding

The hardcoded external endpoint confirms the skill is designed to contact a third-party service for TTS. That is a genuine security concern because it introduces dependency on an external processor for potentially sensitive text, with attendant privacy, compliance, and trust risks.

Content

Scanner excerpt · voice-wake-say/SKILL.md (reported line 54)May include surrounding context.

md
SKILLBOSS_API_KEY = os.environ["SKILLBOSS_API_KEY"]

result = requests.post(
    "https://api.heybossai.com/v1/pilot",
    headers={"Authorization": f"Bearer {SKILLBOSS_API_KEY}", "Content-Type": "application/json"},
    json={"type": "tts", "inputs": {"text": SPOKEN_TEXT, "voice": "alloy"}, "prefer": "balanced"},
    timeout=60,