The Agent Skills Directory

[PROMPT_INJECTION] (SAFE): The skill uses phrases like 'CRITICAL' and 'HIGHEST PRIORITY' to define routing logic and prioritize specific Rust-related topics. These instructions are part of the skill's operational design and do not attempt to override the agent's safety protocols or core instructions.\n- [DATA_EXFILTRATION] (SAFE): The analysis found no code patterns related to network operations, data transmission, or access to sensitive credentials and file paths.\n- [REMOTE_CODE_EXECUTION] (SAFE): The skill does not perform any remote downloads or execution of external scripts. References to tools like cargo audit or miri are provided as suggested commands for user-initiated local audits and are part of standard Rust development workflows.\n- [INDIRECT_PROMPT_INJECTION] (SAFE): Although the skill is designed to process external Rust source code (via globs for Cargo.toml and *.rs), it functions as a diagnostic and reasoning tool without the capability to write files, make network requests, or execute arbitrary code. This limits its role to an informational 'summarizer/reasoner' tier with negligible risk.

rust-router