The Agent Skills Directory

[PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection because it reads and processes data that could be controlled by an external source (e.g., file content via git diff).
Ingestion points: SKILL.md (Processes data from git diff, git status, and the user-supplied argument [description of the changes]).
Boundary markers: Absent. The skill does not instruct the agent to use specific delimiters or ignore instructions embedded in the file content during the automated PR creation process.
Capability inventory: SKILL.md (Capabilities include file staging, committing changes, and creating pull requests via the git and gh tools).
Sanitization: Absent. There is no mention of sanitizing or escaping the data before it is used to construct commit messages or pull request titles.

create-pr