The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses several shell commands to interact with the local git environment, including git status, git log, git diff, and reading files from the .git directory. These are standard operations for a developer tool intended to analyze version control state.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the git history (commit messages and file diffs). While this presents a surface for indirect prompt injection if an attacker crafts malicious commit content, the risk is inherent to the tool's primary purpose of reviewing code changes. There are no specific instructions to execute instructions found within the data.
Ingestion points: git log outputs, git diff outputs, and internal git metadata files (e.g., .git/rebase-merge/onto).
Boundary markers: Not explicitly defined in the prompt instructions.
Capability inventory: Uses bash shell commands to execute git operations.
Sanitization: No specific sanitization or escaping of the git data is performed before it is processed by the agent.

review-conflict