The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute git commands (status, diff, commit) and filesystem utilities (ls, find, grep). This is consistent with its primary purpose of managing version control changes.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process instructions from local and project-based configuration files (.claude/standards/). This creates a potential surface for indirect prompt injection where malicious content in a repository's configuration could attempt to influence agent behavior. The risk is mitigated by explicit instructions to avoid sensitive data commitment and requirements for the agent to report the sources it has read.
Ingestion points: The skill reads files from .claude/standards/ and ~/.claude/standards/ using Read, Glob, and Grep tools.
Boundary markers: The agent is instructed to state the source file path in its output, but there are no technical delimiters or sanitization steps for the file content itself.
Capability inventory: The agent can execute git commands and arbitrary Bash scripts via allowed tools.
Sanitization: No automated sanitization of ingested configuration content is specified.
[EXTERNAL_DOWNLOADS]: The skill metadata references the developer's GitHub repository and an external specification domain (agentskills.io). These resources are associated with the skill's author and ecosystem, representing standard update and documentation paths.

commit