The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, specifically git commit history and file content during conflict resolution, creating a surface for indirect prompt injection attacks.
Ingestion points: SKILL.md (instructions to process git status --porcelain and commit history).
Boundary markers: Absent; there are no specific delimiters to distinguish between code content and potential instructions embedded in commits.
Capability inventory: Shell command execution (git operations).
Sanitization: Absent; the skill does not explicitly filter or validate the content of git logs or files before the agent processes them.

git-sync