imagegen-frontend-web
Pass
Audited by Gen Agent Trust Hub on Jul 16, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill uses assertive directives such as 'HARD OUTPUT RULE' and 'overrides any model default' to strictly enforce a multi-image output format and specific design compositions, which can be used to bypass default model behavior.
- [PROMPT_INJECTION]: An indirect prompt injection vulnerability surface is present as the skill ingests untrusted user input ('brief') and uses it to generate file summaries and logs.
- Ingestion points: The user's design brief or project description provided at runtime.
- Boundary markers: No delimiters or 'ignore' instructions are defined to separate the user-supplied data from the reporting logic.
- Capability inventory: The skill instructs the agent to perform file system operations, including creating files in
llmwiki/wiki/draft/uiux/and appending tollmwiki/wiki/index.mdandllmwiki/wiki/log.md. - Sanitization: There are no instructions to sanitize, validate, or escape the user-influenced summary string ('ten') before it is incorporated into file paths or log entries, potentially allowing for path traversal or log poisoning if the agent does not apply its own filters.
Audit Metadata