execute-planning-subtask

Pass

Audited by Gen Agent Trust Hub on Apr 8, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands using the agkan CLI to manage task lifecycles, including retrieval, updating, and metadata modification.
  • [PROMPT_INJECTION]: The skill processes untrusted task content retrieved from an external CLI, presenting a surface for indirect prompt injection.
  • Ingestion points: External task data retrieved via 'agkan task get --json' in SKILL.md.
  • Boundary markers: Absent; task content is interpolated into update commands without explicit delimiters or instructions to ignore embedded commands.
  • Capability inventory: The skill can add, update, and modify task metadata and descriptions via the agkan CLI.
  • Sanitization: Absent; the skill appends and organizes task content as-is without validation or escaping.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 8, 2026, 06:43 AM