The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill utilizes dynamic imports within scripts/setup-modules.mjs to load functionality from a sibling skill named lodestar-setup. This modular design facilitates code sharing between the author's related skills within the agent's environment.
[COMMAND_EXECUTION]: The skill automates repository-level tasks, including git operations and build/test scripts identified in the project configuration. It maintains safety by requiring user confirmation before committing changes and by using structured command arguments instead of raw shell execution, which reduces the risk of command injection.
[INDIRECT_PROMPT_INJECTION]: The skill processes plan documents and repository configuration files that influence its execution flow. It mitigates potential injection risks by defining clear execution stages, decision gates that pause for user input, and strict boundaries regarding which files the agent is permitted to modify.

lodestar-implement