building-communication-engine

Pass

Audited by Gen Agent Trust Hub on Mar 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of high-level instructional content for business strategy. It facilitates data management by reading local context files and generating markdown reports in a local 'outputs' directory. These operations are standard for its intended use case.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from project context files. While this represents a potential injection surface, the risk is negligible as the skill lacks high-privilege capabilities such as arbitrary command execution or network data exfiltration.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 26, 2026, 04:49 PM
Security Audit — agent-trust-hub — building-communication-engine