modular-monolith-architecture
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied arguments to define project and module names, which are then interpolated into generation prompts to create directory structures and source code. This creates an attack surface where a user could provide strings containing malicious instructions to override the agent's behavior during the scaffolding process.
- Ingestion points: The
$ARGUMENTSvariable inSKILL.mdcaptures user input for the language/framework, project name, and multiple business module names. - Boundary markers: The skill lacks explicit boundary markers or instructions to treat user-provided names as inert data, which could allow the agent to interpret them as instructions.
- Capability inventory: The agent has the capability to write multiple files and create complex directory hierarchies based on these inputs, which could be misused to write unauthorized content if the agent is influenced by the input.
- Sanitization: No validation or sanitization rules are provided to ensure module or project names do not contain prompt injection payloads or control characters.
Audit Metadata