The Agent Skills Directory

[SAFE]: Procedural Workflow. The skill defines a step-by-step process for loading, reviewing, and executing implementation plans, relying on natural language instructions rather than embedded scripts.
[SAFE]: Safety Constraints. The skill includes explicit instructions to avoid modifying master/main branches without consent and mandates stopping the process when blockers or ambiguities are encountered to prevent erroneous actions.
[SAFE]: Modular Integration. The skill utilizes a modular architecture by referencing other skills in the 'superpowers' namespace (e.g., using-git-worktrees, writing-plans, finishing-a-development-branch) to handle specialized tasks and environment isolation.
[SAFE]: Data Ingestion Surface. The skill reads external plan files, which is a common pattern for development tools. It mitigates potential indirect prompt injection by instructing the agent to 'Review critically' and stop if gaps or unclear instructions are identified.

executing-plans