Skills
Navigate
skills/jeremylongshore/tons-of-skills-marketplace/vastai-enterprise-rbac/Gen Agent Trust Hub

vastai-enterprise-rbac

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill includes a function team_spending_report that parses a local JSONL file (vast_audit.jsonl). If an attacker were able to write malicious content to this log file, they could theoretically attempt to influence the agent's summary of the report.
  • Ingestion points: team_spending_report in SKILL.md reads from vast_audit.jsonl.
  • Boundary markers: None; uses standard json.loads processing.
  • Capability inventory: The skill uses Write to generate logs and restricted Bash(vastai:*) commands.
  • Sanitization: Standard JSON parsing is used; the script only extracts numerical total_cost and string team values, minimizing the risk of instructions being executed.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 11:07 AM