event-comms-channels
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is purely instructional and does not contain any executable scripts, shell commands, or network operations.\n- [METADATA_POISONING]: Metadata fields (name, description, author, version) are used appropriately for their intended purpose without deceptive or malicious instructions.\n- [INDIRECT_PROMPT_INJECTION]: The skill features an interview workflow that ingests untrusted user data (Questions 1-10 in SKILL.md). Analysis shows:\n
- Ingestion points: User answers provided during the Interview phase.\n
- Boundary markers: None identified.\n
- Capability inventory: No subprocess calls, exec/eval, file-write, or network operations exist in the skill scripts or instructions.\n
- Sanitization: None identified, but no sinks are present for exploitation.\n The lack of dangerous capabilities renders the indirect prompt injection surface safe.\n- [EXTERNAL_DOWNLOADS]: The skill references well-known services (Discord, Slack, Major League Hacking) and other skills from the same author ('samber'). These references are neutral, documented for functionality, and do not involve remote code execution or data exfiltration.
Audit Metadata