The Agent Skills Directory

[SAFE]: The skill operates entirely within the local project environment, managing documentation and architectural artifacts. It does not utilize network tools, system-level commands, or access sensitive configuration files like credentials or environment variables.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it ingests and processes content from external markdown files (GDDs, ADRs, and Epic manifests) to generate story descriptions and test cases. However, given the skill's restricted capabilities and intended use in a developer workflow, this represents a standard operational surface rather than a critical vulnerability.
Ingestion points: design/gdd/*.md, docs/architecture/*.md, production/epics/*.md, and production/qa/qa-plan-*.md.
Boundary markers: None identified; data is processed as plain text and interpolated into templates.
Capability inventory: Write (creation of project markdown files), Task (delegation of review tasks to an internal qa-lead agent).
Sanitization: None identified; the skill assumes the provided project documentation is well-formed.
[COMMAND_EXECUTION]: The skill uses the Task tool to invoke a sub-agent (qa-lead) using defined gates. This is a controlled, high-level interaction between agent roles and does not involve the execution of arbitrary shell commands or system-level processes.

create-stories