The Agent Skills Directory

[PROMPT_INJECTION]: The skill ingests untrusted data from the local project environment (files, docs, commits) to understand the project context. While it lacks explicit boundary markers or sanitization for this data, its capabilities are restricted to generating dialogue and writing documentation to the local file system, presenting a low surface for indirect prompt injection.
[COMMAND_EXECUTION]: The skill directs the agent to interact with the file system and version control using standard git commands to document designs and manage workspaces. These operations are routine for development workflows and are performed within the local project context.

brainstorming