The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests sprint task descriptions as data for the implementation stage. Since the skill uses the Bash tool to execute these tasks, there is a potential surface for indirect prompt injection where instructions embedded in task data could attempt to manipulate agent behavior. However, the enforced workflow requires a verdict: pass from a review stage before proceeding, which serves as a mitigation gate.- Ingestion points: Sprint task definitions (SKILL.md).- Boundary markers: None explicitly defined to isolate task content from instructions.- Capability inventory: Bash, Read, Write tools.- Sanitization: No sanitization or escaping mechanisms for external task content are described.

kyro-sprint-executor