The Agent Skills Directory

[DYNAMIC_CONTEXT_INJECTION]: The skill utilizes the !bash syntax in the markdown body to execute a shell script (yaml-helper.sh) at the time the skill is loaded. This execution is used to resolve configuration keys related to automation and workflow tiers.
[COMMAND_EXECUTION]: The skill executes a local script located at ${CLAUDE_SKILL_DIR}/../../hooks/yaml-helper.sh. This script is used both for pre-load configuration resolution and within the allowed-tools scoping to manage YAML-based settings.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it ingests data from external files, including story files (production/epics/**/story-*.md) and Game Design Documents (GDDs). Malicious instructions embedded in these files could theoretically influence the agent's output during the plan generation phase.
Ingestion points: Story files (production/epics/), GDD files (design/gdd/), and sprint plans (production/sprints/).
Boundary markers: None explicitly defined for isolating the content of the Acceptance Criteria or Engine notes sections during processing.
Capability inventory: The skill has access to Write, Edit, and limited Bash capabilities, which are used to generate the final QA plan and back-fill test cases into story files.
Sanitization: No specific sanitization or escaping of external content is mentioned before it is interpolated into the test plan templates.

qa-plan