lld-writer

Pass

Audited by Gen Agent Trust Hub on Mar 8, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No direct prompt injection or safety bypass instructions were found. The skill does process external data (design documents), which presents an indirect injection surface; however, this is inherent to its primary purpose.
  • Ingestion points: Reads local files including PRD, HLD, Contract, Guardrails, and ADR.
  • Boundary markers: Not specified in the current instructions.
  • Capability inventory: Limited to text generation and progress tracking via local tools.
  • Sanitization: Not explicitly mentioned.
  • [DATA_EXFILTRATION]: The skill accesses local design files for context as intended. There are no network operations, external URLs, or hardcoded credentials detected that would facilitate exfiltration.
  • [COMMAND_EXECUTION]: The skill uses local tracking (TodoWrite) and file discovery (Glob) mechanisms. It does not attempt to execute arbitrary shell commands, script downloads, or privilege escalation.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 8, 2026, 12:16 AM