The Agent Skills Directory

[SAFE]: The skill implements a legitimate developer workflow for code review. It uses established git commands to compare commits and generates a report based on standard engineering criteria.
[COMMAND_EXECUTION]: The skill uses local git commands (git rev-parse, git log, git diff) to identify changes. These are standard operations for development tools and do not involve privilege escalation or arbitrary execution of untrusted input.
[DATA_EXFILTRATION]: No network operations or external data transfers were identified. The code changes are processed by a subagent within the local environment.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as the reviewer subagent processes untrusted code content.
Ingestion points: The code-reviewer.md template ingests code changes via git diff output.
Boundary markers: The prompt template lacks explicit delimiters or instructions to ignore embedded commands within the diff content.
Capability inventory: The subagent uses a general-purpose task tool, which may have access to local file systems or shell commands depending on the platform environment.
Sanitization: No sanitization or escaping of the diff content is performed before it is interpolated into the reviewer's prompt.

requesting-code-review