The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands including cargo fmt, cargo clippy, and cargo test. Running cargo test involves executing the test suite of the Rust project being analyzed, which can result in arbitrary code execution on the host environment if the target project is malicious.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted Rust source code while maintaining command execution capabilities, creating an attack surface for indirect prompt injection.
Ingestion points: The agent reads user-provided Rust files, modules, crates, and diffs as specified in the "Decide the scope" section of SKILL.md.
Boundary markers: There are no instructions to use boundary markers or delimiters when reading the target code, nor warnings to ignore embedded instructions within the code comments.
Capability inventory: The skill uses the Bash tool to run the Rust toolchain, specifically cargo test, which executes code from the project root.
Sanitization: No validation or sanitization of the input code is performed before execution via cargo tools.
[EXTERNAL_DOWNLOADS]: The skill recommends that users adopt lint configurations from a remote repository at https://github.com/morganlinton/BeautifulRust. This repository is owned by the skill author and is a legitimate vendor resource used to provide configuration for the review process.

rust-review