The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the ingestion of external project artifacts.
Ingestion points: The skill reads and interprets content from external files such as 00-requirements.md, 02-to-be-plan.md, and 03-implementation-summary.md (found in SKILL.md).
Boundary markers: There are no explicit delimiters or 'ignore embedded instructions' warnings defined to separate user-provided data from the agent's core instructions.
Capability inventory: The skill utilizes the Task tool to create and dispatch subagents based on the content of these external files.
Sanitization: The skill interpolates external content into subagent prompts without any evidence of validation or sanitization, allowing potentially malicious instructions in the project files to influence subagent behavior.

rlm-subagent