Back to skill

Security audit

toddler-meltdown-decoder

Security checks for vulnerabilities and agentic risk

Overview

The skill’s parenting guidance and local episode logging fit its purpose, with privacy precautions needed for sensitive child records.

Use triage without logging if you do not want persistent records. Logs can contain sensitive child information: avoid identifying details, restrict access to scripts/data/meltdowns.json, review caregiver handoffs, and delete records when no longer needed. Treat scores as parenting prompts rather than a diagnosis, and follow professional medical guidance.

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
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (4)

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_read, file_write.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The skill instructs users to record sensitive child health and behavioral information without a privacy warning.

Medium
Category
Output Handling
Confidence
88% confidence
Finding

The skill encourages persistent recording of a young child's age, behavioral cues, possible health symptoms, and free-text notes without explaining storage, retention, or privacy safeguards. Such records could expose sensitive information through shared devices, backups, or caregiver handoffs; the child-focused context increases sensitivity, although no exfiltration or insecure storage implementation is demonstrated.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · references/de-escalation-playbook.md (reported line 80)May include surrounding context.

md
- Now: check (fingers in mouth, ears, fever); offer cold chew
  thing; appropriate weight-based analgesic per your
  pediatrician's dosing if indicated.
- Never: assume behavior is "just a phase" for 3+ days without
  checking — ear infections present as tantrums.
- Prevent: log night-wakings alongside meltdowns; the
  correlation is your early-warning system.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · scripts/meltdown_triage.py (reported line 112)May include surrounding context.

python
"PREVENT: daily 10-min named special time, child-led."),
    "pain": ("NOW: check mouth/ears/fever; cold chew thing; weight-based "
             "analgesic per pediatrician dosing if indicated.\n"
             "NEVER: assume 'phase' 3+ days without checking - ear "
             "infections present as tantrums.\n"
             "PREVENT: log night-wakings next to meltdowns."),
}