The Agent Skills Directory

[COMMAND_EXECUTION]: The skill frequently executes local Python scripts using uv run. These scripts, such as resolve_customization.py and resolve_config.py, are located in the project's _bmad/scripts/ directory.
[COMMAND_EXECUTION]: The workflow involves automated browser exploration using playwright-cli to open target URLs, capture snapshots, and take screenshots. These operations are performed via shell command execution.
[DYNAMIC_EXECUTION]: The skill utilizes a post-completion hook that resolves a command from the workflow.on_complete configuration key and executes it directly in the terminal. This allows for the execution of arbitrary code defined in project-level .toml files.
[EXTERNAL_DOWNLOADS]: The agent is instructed to install additional components from the vendor's repository (bmad-code-org/bmad-method-test-architecture-enterprise) using npx skills add if they are not already present.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, including product requirement documents (PRDs), architecture decision records (ADRs), and user stories. This data influences the generated test plans and browser exploration targets. There are no explicit sanitization procedures or boundary markers implemented for these inputs.

bmad-testarch-test-design