skills/owainlewis/blueprint/multitask/Gen Agent Trust Hub

multitask

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were identified in the skill instructions.
  • [COMMAND_EXECUTION]: The skill utilizes git commands for branch and worktree management to isolate parallel tasks.
  • [PROMPT_INJECTION]: The skill defines an attack surface for indirect prompt injection via ticket data ingestion.
  • Ingestion points: Ticket descriptions and arguments in SKILL.md.
  • Boundary markers: Absent; ticket content is interpolated into worker prompts without explicit delimiters.
  • Capability inventory: Subagents have full sessions for code implementation, running checks, and PR creation.
  • Sanitization: No sanitization or validation of ticket content is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 08:34 AM
Security Audit — agent-trust-hub — multitask