The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it is designed to ingest and analyze untrusted external data, such as existing game project files, code, and design documents.
Ingestion points: External project files and codebases are processed through commands like bmad-gds-document-project and bmad-gds-dev-story (referenced in SKILL.md and REFERENCE.md).
Boundary markers: The instructions do not specify any delimiters or safety markers for isolating external data from the agent's core instructions.
Capability inventory: The skill utilizes high-privilege tools including Bash, Write, and Read to perform its primary development and testing tasks.
Sanitization: No explicit sanitization or validation of the ingested project data is described in the skill's instructions.
[EXTERNAL_DOWNLOADS]: The skill documentation includes references to external GitHub repositories for installation and further reference.
Evidence: References to github.com/supercent-io/skills-template for installation and github.com/bmad-code-org/bmad-module-game-dev-studio for documentation are neutral references to a well-known service provider.

bmad-gds