The Agent Skills Directory

[SAFE]: The skill outlines a legitimate code review process. It uses standard git and sc (source control) commands to inspect worktrees, repository state, and code diffs.\n- [SAFE]: No evidence of prompt injection, data exfiltration, obfuscation, or unauthorized command execution was found in the skill instructions or metadata.\n- [INDIRECT_PROMPT_INJECTION]: The skill has a potential attack surface as it is designed to process untrusted code changes from external sources. However, this is inherent to its primary purpose as a code review tool.\n
Ingestion points: SKILL.md indicates the agent reads git diffs, worker result artifacts, and Workbench plans.\n
Boundary markers: No specific delimiters or instructions to ignore embedded instructions are defined for the ingested code content.\n
Capability inventory: The agent has access to run_workspace, write_artifact, git, and the sc command suite.\n
Sanitization: No explicit sanitization or filtering of the input code is described; the skill relies on the agent's internal safety guardrails.

review