The Agent Skills Directory

Prompt Injection (LOW): The skill is susceptible to indirect prompt injection (Category 8). It processes task schedules from other skills and interpolates that data into agent prompts.
Ingestion points: The executeTask function ingests ScheduledTask objects from an external schedule.
Boundary markers: No explicit delimiters or boundary markers (like XML tags) are shown in the buildPromptForNode usage to isolate untrusted task content from the agent's instructions.
Capability inventory: The skill uses the Task tool, which has the power to spawn new agents and execute additional tasks. It also has Write and Edit permissions.
Sanitization: There is no evidence of sanitization or validation of the nodeId or description fields before they are passed to the sub-agent's prompt or used to determine the subagent_type.

dag-parallel-executor