The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized network operations were detected. The skill follows established software development best practices.
[COMMAND_EXECUTION]: The skill utilizes common shell commands such as mkdir, diff, and git to manage local file structures and version control branches. These operations are standard for development automation and occur within the local environment.
[INDIRECT_PROMPT_INJECTION]: The skill manages a workflow where subagents process a design document (docs/plans/<feature>/design.md). While this presents a potential ingestion point for untrusted instructions if the document is externally sourced, it is a core functional requirement of the tool's intended purpose.
Ingestion points: Implementation teams read the design document during Phase 3.
Boundary markers: No explicit delimiters are used for the design content, though instructions are provided to agents to guide their planning process.
Capability inventory: Subagents are equipped with writing-plans and executing-plans tools for file modification and command execution within dedicated worktrees.
Sanitization: The skill does not perform specific sanitization of the design document's text.

cookoff