The Agent Skills Directory

[SAFE]: The skill operates within the expected scope of a development assistant, focusing on local file management and user interaction.
[COMMAND_EXECUTION]: The skill instructs the agent to run 'yarn prep' to verify code quality. This is a common and legitimate practice in software development workflows.
[EXTERNAL_DOWNLOADS]: The skill contains references to 'https://codely.com', which is the official website of the skill's author (Codely). These links are used for documentation and attribution purposes.
[PROMPT_INJECTION]: The skill demonstrates an inherent surface for indirect prompt injection as it reads repository files to generate context.
Ingestion points: Accesses 'AGENTS.md' and repository source files in the 'docs/' directory.
Boundary markers: None identified in the prompt instructions for separating file content from agent instructions.
Capability inventory: File system access (write to '.agents/plans/'), shell command execution ('yarn prep'), and invocation of other agent skills ('/git-commit', '/plan-implement-phase').
Sanitization: Content is processed as-is to ensure architectural consistency with the existing codebase.

plan-create