execute-task

Pass

Audited by Gen Agent Trust Hub on Feb 25, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: Susceptibility to Indirect Prompt Injection.
  • Ingestion points: Reads from 'specs/tasks.md', 'specs/product_specs.md', and 'specs/architecture.md'.
  • Boundary markers: Uses labels like 'GOAL:' and 'ACTION ITEMS:' but lacks strict delimiters or instructions for subagents to ignore nested commands.
  • Capability inventory: The 'task-executor' subagent possesses 'implement-feature' and 'use-git-worktree' skills, allowing file and repository modifications.
  • Sanitization: Content is interpolated directly into agent prompts without escaping or validation.
  • [COMMAND_EXECUTION]: Executes shell and git commands.
  • Uses 'git worktree', 'git merge', and 'git branch' to manage code changes.
  • These actions are fundamental to the skill's purpose of orchestrating task implementation.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 25, 2026, 09:41 PM