The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands and local scripts to manage the development environment and TDD cycle.\n
Evidence: Uses cat, grep, and pip show to verify package versions in configuration files.\n
Evidence: Runs local scripts such as read_file.py and run_tests.py from the referenced bmad-commands library.\n- [EXTERNAL_DOWNLOADS]: The skill accesses official documentation sources via the internet.\n
Evidence: Uses WebFetch to retrieve information from well-known services like npmjs.com, pypi.org, pkg.go.dev, and readthedocs.io to ensure API compatibility.\n- [PROMPT_INJECTION]: The skill processes untrusted input from task files and the web, creating a potential surface for indirect prompt injection.\n
Ingestion points: Loads content from workspace/tasks/{task_id}.md and web-based documentation.\n
Boundary markers: No explicit markers or 'ignore' instructions are provided to isolate processed data from agent instructions.\n
Capability inventory: Possesses file system write access and the ability to execute commands and scripts.\n
Sanitization: Does not perform validation or sanitization on the external content before it is used to drive the TDD cycle.

implement