backend-development-feature-development

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill facilitates indirect prompt injection by interpolating untrusted user input into multiple sub-agent execution steps across all development phases.
  • Ingestion points: The $ARGUMENTS variable in SKILL.md (Steps 1 through 12) ingests user-supplied feature names and descriptions into the context of specialized sub-agents.
  • Boundary markers: There are no delimiters (such as XML tags or triple backticks) or explicit instructions to ignore embedded commands surrounding the interpolated user input. This allows a malicious user to craft a feature description that could hijack the logic of the sub-agents.
  • Capability inventory: The workflow spawns sub-agents with high-impact capabilities, including backend and frontend code generation, security auditing, CI/CD pipeline configuration, and data engineering.
  • Sanitization: The skill lacks any visible mechanisms for escaping, validating, or filtering the $ARGUMENTS content before it is processed by the agentic pipeline.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 04:34 PM
Security Audit — agent-trust-hub — backend-development-feature-development