The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from external project files.
Ingestion points: Document files such as docs/prd-*.md, docs/tech-spec-*.md, and docs/architecture-*.md are read into the agent's context during the 'Extract and Inventory' phase in commands/sprint-planning.md.
Boundary markers: The instructions do not define delimiters or provide warnings to the agent to ignore instructions embedded within the ingested requirement documents.
Capability inventory: The skill utilizes Read, Edit, Write, and TodoWrite tools to perform operations on the local file system and manage progress based on the processed inputs.
Sanitization: There is no evidence of validation or sanitization logic to strip potential malicious instructions from the requirements documents before they influence the agent's story creation and planning tasks.

scrum-master