22-personal-brand-context
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection because it ingests untrusted user input and persists it into a file (
.agents/personal-brand-context.md) intended for consumption by other AI agents (Skills 23-28). - Ingestion points: User responses during the interview process in
SKILL.md(Step 3: Thu thap thong tin theo variant). - Boundary markers: The skill uses Markdown headers and tables to structure data but does not include explicit delimiters or 'ignore instructions' warnings for downstream agents.
- Capability inventory: Performs local file-write operations within the project's
.agentsdirectory. - Sanitization: No explicit sanitization or validation of the user's input strings is defined in the instructions.
- [DATA_EXPOSURE]: The skill creates a local file containing user-provided professional and personal details. However, it does not attempt to access system secrets, SSH keys, or environment variables, and it performs no network operations to exfiltrate this data.
Audit Metadata