The Agent Skills Directory

[SAFE]: A comprehensive analysis of the skill's instructions and reference guidance revealed no malicious patterns, obfuscation, or unauthorized commands.
[DATA_EXPOSURE]: The skill demonstrates proactive security awareness by including a specific directive in references/current-guidance.md to avoid encoding secrets, credentials, private task text, or customer names within the cycle contracts.
[INDIRECT_PROMPT_INJECTION]: Although the skill is designed to process external inputs such as user requests and specifications, it implements logical safeguards by requiring the definition of input scopes, allowed/forbidden change boundaries, and explicit stop conditions to maintain agent control.

cycle-contract