The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides numerous shell commands for managing Rust projects, including cargo tree for dependency analysis, cargo-deny for policy enforcement, and cargo-audit for security scanning. These commands are standard for Rust development and align with the skill's primary purpose.
[EXTERNAL_DOWNLOADS]: The skill references several external resources and tools for dependency management. These include the official Rust documentation, the RustSec advisory database, and well-known tools like Renovate, Dependabot, and cargo-deny. References to these services are informative and use neutral language to guide the user in setting up secure workflows.
[DATA_EXPOSURE]: The documentation contains instructions on managing sensitive files like ~/.cargo/credentials and using environment variables for tokens. It explicitly advises against committing credentials to version control, promoting secure secret management practices.
[PROMPT_INJECTION]: The skill operates on external data such as Cargo manifests (Cargo.toml), lockfiles, and vulnerability databases. While this presents an indirect prompt injection surface (Category 8), the skill focuses on providing guidance for manual or agent-assisted governance rather than automated execution on untrusted third-party inputs without oversight.
Ingestion points: Cargo.toml, Cargo.lock, deny.toml, and the RustSec advisory database.
Boundary markers: None explicitly specified for the agent's interaction with the data.
Capability inventory: Use of cargo CLI tools and specialized security scanners.
Sanitization: Relies on the standard parsing behavior of the underlying Rust tooling.

rust-dependencies