The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes git and gh (GitHub CLI) to manage branches and pull requests as part of its planning workflow. These are standard engineering tools used as intended for project management.
[INDIRECT_PROMPT_INJECTION]: The skill reads project files and existing plans to gather context for plan generation. This process involves ingesting untrusted data which could potentially contain malicious instructions. However, the skill mitigates this by incorporating an 'adversarial review gate' where a stronger model (e.g., Opus) checks the generated plan against an anti-pattern catalog before finalization.
[NO_CODE]: The skill consists entirely of Markdown instructions and does not include any executable scripts, shell hooks, or binary files. It relies on the agent's native capabilities to perform file and command operations.
[EXTERNAL_DOWNLOADS]: The installation instructions provide guidance for users to verify the skill using standard Git commands (git fetch, git log, git checkout). These are administrative instructions for the user to manage the skill's source code safely from GitHub.

blueprint