The Agent Skills Directory

[SAFE]: The skill consists of instructional documentation and reference guides designed to improve the quality and safety of C code. All code snippets provided are either examples of vulnerabilities to avoid or patterns for remediation.
[PROMPT_INJECTION]: The skill includes an instruction to follow project-specific conventions in files like AGENTS.md or CLAUDE.md and allows them to 'win' over the skill's default rules. While this is a standard pattern for repository-level configuration, it represents an indirect prompt injection surface if an agent operates on a repository containing malicious instructions in those specific configuration files.
[COMMAND_EXECUTION]: The specification documentation (SPEC.md) references local Python scripts (distiller.py) and standard C development tooling (GCC, Clang, Valgrind) used for evaluation and the development workflow. These are standard tools for the skill's domain and are not invoked automatically via dangerous patterns.

c-systems