The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands to execute test runners (pytest, vitest) and perform version control operations (git diff). These commands are standard for development workflows and are used here within their intended scope.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface because it ingests project code and user requirements to generate test plans and scripts.
Ingestion points: External data enters the agent context through file reads of project source code, Git diff output, and user-provided functional descriptions.
Boundary markers: The skill does not explicitly define delimiters for untrusted data, although it uses a structured YAML manifest (test-manifest.yaml) to manage test metadata.
Capability inventory: The skill has the capability to write files (SKILL.md, test scripts) and execute shell commands (pytest, vitest, git).
Sanitization: The risk of malicious code generation is significantly mitigated by a mandatory "human-in-the-loop" process. The agent must generate a test-plan.md for user review and confirmation before any actual test scripts are created or modified.

test-suite-maintainer