The Agent Skills Directory

[COMMAND_EXECUTION]: The orchestrator utilizes the Bash tool throughout multiple phases to perform file discovery, content searching, and metadata analysis via standard system utilities like find, rg, and grep.
[COMMAND_EXECUTION]: In the verification phase (Phase 7), the skill dynamically identifies and executes language-specific test runners such as npm test, pytest, cargo test, and go test to evaluate TDD compliance within the local project environment.
[PROMPT_INJECTION]: The skill processes untrusted data, establishing an indirect prompt injection surface. 1. Ingestion points: User goals collected via AskUserQuestion and local source code ingested via Read and Grep. 2. Boundary markers: Agent prompts utilize Markdown headers and structured templates to separate instructions. 3. Capability inventory: The skill has permission to execute shell commands (Bash), modify files (Write, Edit), and manage tasks. 4. Sanitization: No explicit sanitization or escaping logic is implemented in the skill's orchestration layer.
[EXTERNAL_DOWNLOADS]: The skill references the mcp__exa tool in Phase 5 for fetching TDD best practices. This targets Exa, a well-known technology service for AI agents, and is used for its intended informative purpose.

workflow-tdd-plan-plan