task-splitting

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions are limited to natural language processing for task splitting and do not involve executable code, shell commands, or network access.
  • [DATA_EXFILTRATION]: No sensitive file access or data transmission patterns were identified. The skill only interacts with project documentation (docs/Implementation Plan.md).
  • [PROMPT_INJECTION]: The content does not contain any instructions to ignore previous rules, override safety filters, or disclose system prompts.
  • [REMOTE_CODE_EXECUTION]: There are no patterns indicating the download or execution of external scripts or packages.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a minimal attack surface for indirect injection as it processes user-provided tasks.
  • Ingestion points: Input task description.
  • Boundary markers: None provided in the prompt template.
  • Capability inventory: Limited to writing text to a documentation file (docs/Implementation Plan.md).
  • Sanitization: No specific sanitization or escaping of input data is defined.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 07:32 AM
Security Audit — agent-trust-hub — task-splitting