skills/hijabcantic-dotcom/app-review-opportunity-research-template/app-review-evidence-research/Gen Agent Trust Hub
app-review-evidence-research
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a vulnerability to indirect prompt injection (Category 8) due to the ingestion and processing of untrusted external content.
- Ingestion points: In
SKILL.md, the workflow requires the agent to collect and analyze 20 to 50 recent low-rating reviews from public app stores. - Boundary markers: The instructions do not provide explicit delimiters (like XML tags or triple quotes) or warnings to ignore instructions that might be embedded within the gathered review text.
- Capability inventory: The agent uses standard text processing capabilities to analyze the data, but without boundary markers, an attacker-controlled review could potentially influence the agent's behavior during the analysis phase.
- Sanitization: The skill only instructs the agent to redact personal data for privacy; it does not mention sanitizing the content for malicious prompts or control characters.
Audit Metadata