backend-development-feature-development

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill facilitates an indirect prompt injection surface by interpolating untrusted user input ($ARGUMENTS) into prompts sent to multiple downstream sub-agents.
    • Ingestion points: The --feature argument is ingested via the $ARGUMENTS placeholder and passed to sub-agents in Phase 1 (Discovery), Phase 2 (Implementation), Phase 3 (Testing), and Phase 4 (Deployment).
    • Boundary markers: The instructions lack delimiters or specific system prompts to isolate the user-provided feature description from the agent's instructions, potentially allowing the input to influence the sub-agents' behavior.
    • Capability inventory: The targeted sub-agents (e.g., backend-architect, deployment-engineer, test-automator) have the capability to generate source code, define API contracts, and configure CI/CD pipelines.
    • Sanitization: There is no evidence of input validation, escaping, or filtering to prevent malicious instructions within the feature description from being interpreted by the sub-agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 11:40 AM
Security Audit — agent-trust-hub — backend-development-feature-development