The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script {project-root}/_bmad/scripts/resolve_customization.py to resolve configuration blocks from base, team, and user-level TOML files.
[DYNAMIC_EXECUTION]: The workflow processes and executes arbitrary steps defined in the activation_steps_prepend, activation_steps_append, and on_complete configuration keys, allowing for runtime injection of instructions or commands based on local file content.
[INDIRECT_PROMPT_INJECTION]: The skill has a high surface area for indirect injection due to its automated ingestion of project data.
Ingestion points: The skill loads project facts from config.yaml and all instances of project-context.md found within the project root, as well as reading various test files and results during its metrics gathering phase.
Boundary markers: The instructions do not define strict delimiters or instructions to the agent to disregard commands embedded within the ingested files.
Capability inventory: The skill can execute Python scripts and write review reports to the project's output directory.
Sanitization: There is no logic present to sanitize or validate the content of the external markdown or YAML files before they are added to the foundational context.

gds-test-review