The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's activation sequence in SKILL.md involves executing a Python script (resolve_customization.py) via the uv run command. This is used to resolve and merge agent customization settings from the project environment.
[DYNAMIC_EXECUTION]: The skill dynamically executes logic from a computed path ({project-root}/_bmad/scripts/resolve_customization.py) during its initialization phase to handle environment-specific overrides.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection through the ingestion of external data during activation.
Ingestion points: Merges customization data from customize.toml and project-level override files (.toml, .user.toml) in SKILL.md Step 1, and loads persistent facts from files or globs specified in {agent.persistent_facts} in Step 4.
Boundary markers: The instructions do not define clear delimiters or "ignore instructions" warnings for the content loaded from these external sources when they are integrated into the agent's active persona and context in Step 3.
Capability inventory: The skill has capabilities for shell command execution (via activating other skills like gds-dev-story), file manipulation, and code review.
Sanitization: No sanitization or validation logic is present for the metadata and facts loaded from the project-controlled files.

gds-agent-game-dev