The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a configuration resolver script located at {project-root}/_bmad/scripts/resolve_customization.py during activation and completion. It also processes automated execution steps defined in the activation_steps_prepend and activation_steps_append arrays within the customization TOML files.
[PROMPT_INJECTION]: The skill populates the agent's context with potentially untrusted data from the project directory, creating a surface for indirect prompt injection.
Ingestion points: The skill loads files matching the glob {project-root}/**/project-context.md into persistent facts and reads test files and execution results from the workspace.
Boundary markers: No specific delimiters or safety instructions are used to separate loaded project content from the core system instructions.
Capability inventory: The agent has the ability to execute shell commands via python3 and write files to the project directory.
Sanitization: Content from the project files is loaded directly into the context without validation or sanitization.

gds-test-review