The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external, user-controlled data from GitHub commit messages, author metadata, and code diffs.
Ingestion points: Commands such as gh api repos/owner/repo-name/commits and git log retrieve content from external sources.
Boundary markers: None present in the provided shell examples. Data is retrieved and piped to jq or standard output without wrapping in delimiters.
Capability inventory: The skill utilizes network API calls via gh and file system operations via git and cd commands.
Sanitization: The skill does not include instructions or scripts to sanitize or filter out potential instructions embedded in commit messages or diff contents.
[COMMAND_EXECUTION]: The skill relies on shell command execution for its primary functionality.
Evidence: Multiple examples using gh api, git log, git show, and git diff to interact with repositories.
Context: These are standard development tools used for their intended purpose in a developer-oriented skill.

commit-operations