full-stack-orchestration-full-stack-feature

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data via the $ARGUMENTS placeholder and interpolates it directly into prompts for high-capability sub-agents without sufficient isolation or sanitization.
  • Ingestion points: User input enters the workflow through the $ARGUMENTS variable, which is used in every phase of the instructions in SKILL.md (Steps 1 through 12).
  • Boundary markers: The skill lacks explicit delimiters or boundary markers (such as triple backticks or XML tags) around the interpolated $ARGUMENTS content, which could allow malicious instructions within the input to be interpreted as authoritative commands by the sub-agents.
  • Capability inventory: The sub-agents triggered by this skill (e.g., deployment-engineer, database-architect, sql-pro) have broad capabilities, including generating infrastructure-as-code, executing database migrations, and configuring CI/CD pipelines.
  • Sanitization: There is no evidence of input validation or sanitization applied to the $ARGUMENTS string before it is passed to the specialized agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 05:38 PM
Security Audit — agent-trust-hub — full-stack-orchestration-full-stack-feature