The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions and workflow involve reviewing and processing user-provided Rust source code, which constitutes an ingestion point for untrusted data.
Ingestion points: The agent ingests code during tasks related to 'coding convention', 'code review', and 'best practice' as defined in the triggers.
Boundary markers: The instructions do not specify boundary markers or warnings to disregard instructions embedded within the code being analyzed.
Capability inventory: The skill recommends executing cargo test, cargo clippy, and cargo doc (documented in SKILL.md, SKILL_EN.md, and SKILL_ZH.md). In the Rust ecosystem, these commands compile and execute code, potentially triggering malicious logic hidden in build.rs files or procedural macros.
Sanitization: There is no mention of sanitizing or validating the safety of external code before running the verification commands.

rust-coding