The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to detect environment configurations and run project tests. Evidence in TEST_COMMANDS.md includes the use of standard CLI tools like jq, grep, and test to inspect local project files such as package.json, Cargo.toml, and pyproject.toml. It also invokes standard test runners like npm test, cargo test, and pytest.
[DYNAMIC_EXECUTION]: The skill identifies and executes commands defined within project metadata. Specifically, TEST_COMMANDS.md facilitates extracting the test script from package.json to be executed by the agent. While this involves running variable command strings, it is a standard behavior for automated developer tools.
[INDIRECT_PROMPT_INJECTION]: The skill processes external requirements documents (PRDs) to generate its output, which represents a potential injection surface.
Ingestion points: The 'Requirements Analysis' workflow in SKILL.md ingests feature descriptions and PRD files.
Boundary markers: The skill does not currently specify the use of delimiters or 'ignore' instructions for the processed requirement data.
Capability inventory: The skill possesses capabilities for shell command execution (via test runners) and file system writes (generating markdown task lists).
Sanitization: No explicit sanitization or validation logic is defined for the input requirement text.

generate-tasks