The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes shell commands using the exec function to perform git add and git commit operations. While these are used for tracking review progress, direct shell execution is a sensitive capability.
[INDIRECT_PROMPT_INJECTION]: The skill reviews external code files by interpolating their content directly into prompts for sub-agents. This creates an attack surface where malicious instructions inside a codebase could manipulate the review outcome or sub-agent behavior.
Ingestion points: The files list containing content from changed source code files is passed to the reviewer agents via the Task prompt template.
Boundary markers: The prompt template does not use specific delimiters or instructions to the LLM to ignore potentially malicious embedded instructions within the provided file content.
Capability inventory: The skill has access to spawn parallel agents (Task), modify files (Edit), and execute shell commands (exec).
Sanitization: There is no evidence of sanitization, filtering, or escaping of the file content before it is processed by the sub-agents.

orchestrate-review