The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is granted access to the Bash tool to facilitate code implementation and test execution. This capability allows for arbitrary command execution within the agent's environment, which is a necessary but high-impact permission for the intended TDD workflow.
[PROMPT_INJECTION]: The workflow is susceptible to indirect prompt injection due to the ingestion of workspace data without explicit sanitization or instruction boundaries.
Ingestion points: Local source code and search results are retrieved via mcp__julie__fast_search, mcp__julie__get_symbols, and the Read tool (SKILL.md).
Boundary markers: There are no instructions to use delimiters or ignore potential directives embedded within the code or files being processed.
Capability inventory: The skill has broad capabilities to modify the environment via Bash, Edit, and Write tools (SKILL.md).
Sanitization: The skill lacks logic to sanitize or validate external content before it is processed by the agent.

tdd-powerhouse