T09 · Insecure Skill Coding Practices
Warning
- Location
scripts/medication_commander.py:137- Finding
Insecure Plaintext Storage of Medication Adherence Data
- Content
View full analysis
- Remediation
View remediation
Security audit
medication-commander
Security checks for vulnerabilities and agentic risk
Overview
This skill is a local medication organizer whose sensitive adherence storage is disclosed and purpose-aligned, but users should understand it stores health-related records in plaintext.
Install only if you are comfortable keeping medication adherence history in a local plaintext file under your home directory. On shared, backed-up, or synced machines, consider restricting file permissions, deleting the adherence file when no longer needed, or avoiding the adherence feature.
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 (1)
Vulnerability Patterns
- 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
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
- Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)
Undeclared Tool Scope
Medium
- Category
- MCP Least Privilege
- Confidence
- 91% confidence
- Finding
The skill documentation exposes file read/write behavior via the adherence feature storing data in a user home directory, but it does not declare any explicit tool scope such as permissions or allowed-tools. In an agent ecosystem, undeclared filesystem access weakens least-privilege guarantees and can let the skill read or modify files beyond what a reviewer or runtime policy expects.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The documentation explicitly states that medication adherence records are stored persistently in a predictable file under the user's home directory, but provides no warning that this data is sensitive health information. In the context of a medication-management skill, these records can reveal medical conditions and treatment patterns, increasing privacy risk if the file is exposed through shared accounts, backups, logs, or weak local permissions.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 94% confidence
- Finding
The script persists adherence history to a predictable file in the user's home directory containing medication names, dosing times, and taken/missed status, which is sensitive health information. In a medication-management context this is especially sensitive, and the code provides no notice, consent flow, permission hardening, or retention controls, increasing the chance of unintended local disclosure on shared systems or through backups/sync.
- Content
Missing User Warnings
Low
- Category
- Not specified by scanner
- Confidence
- 83% confidence
- Finding
The script loads user-supplied JSON medication lists, which inherently contain sensitive personal health data, but the usage text and argument help do not warn users about handling or exposing this information. For a medication-management tool, this is a meaningful privacy disclosure gap in the user-facing interface.
- Content
Static analysis
No suspicious patterns detected.