The Agent Skills Directory

[PROMPT_INJECTION]: Indirect prompt injection surface identified as the skill processes untrusted repository data.
Ingestion points: The workflow involves reading source files, configurations, and specifications during context establishment and scope inventory.
Boundary markers: The instructions do not provide delimiters or specific guidelines to ignore instructions that may be embedded within the files being reviewed.
Capability inventory: The skill utilizes subprocess capabilities by directing the agent to run commands like git status, git rev-parse, rg, cargo metadata, make, and cargo doc.
Sanitization: There is no logic provided to sanitize or validate the content retrieved from the files before it is processed by the agent.
[COMMAND_EXECUTION]: The skill relies on system-level command execution to perform its primary function.
Evidence: The workflow explicitly lists commands such as git, make, cargo, and rg to gather evidence and establish the review context.

review-core