technical-writer
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHCOMMAND_EXECUTION
Full Analysis
- Privilege Escalation (HIGH): In the 'Troubleshooting' section, the skill provides a template for resolving permission issues using 'sudo chown -R $(whoami) ~/.npm'. Encouraging the use of superuser privileges to modify directory ownership is a dangerous pattern that can be exploited if an agent executes these commands or advises users to do so without proper context.
- Indirect Prompt Injection (LOW): The skill acts as a template engine that ingests external project data (ingestion points in SKILL.md templates). It lacks boundary markers or sanitization logic, creating an attack surface where untrusted data could be used to manipulate the generated output (Category 8).
- Unverifiable Dependencies & Remote Code Execution (LOW): The skill's templates include commands for package installation ('npm install', 'yarn add') and remote API calls ('curl'). While these are provided as instructional examples, they constitute a surface for potential command execution if the agent is directed to implement the documentation it generates.
Recommendations
- AI detected serious security threats
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata