The Agent Skills Directory

[Secure Coding Guidelines]: The skill mandates using Result-based error handling instead of panic!, unwrap(), or expect() in library code. This is a standard stability practice that prevents malformed input from causing process termination, which is particularly important in embedded environments like FFI, WASM, or Node.js.
[Input Sanitization and Presentation]: The instructions establish a clear boundary between terminal-specific styling (ANSI colors) and machine-readable data. By restricting color codes to the entry point and omitting them from JSON or host channels, the skill prevents potential injection or corruption in logs and automated monitoring systems.
[Stable Diagnostic APIs]: By requiring stable, machine-readable error codes, the skill enables AI agents and developer tools to programmatically handle errors. This reduces reliance on unstructured string parsing, which can be a source of logic errors or secondary vulnerabilities.
[Algorithmic Robustness]: The guide specifically discourages the use of recursion and regular expressions in diagnostic logic, such as source location scanning. This proactive measure mitigates the risk of stack overflow or Regular Expression Denial of Service (ReDoS) attacks when processing potentially untrusted developer input.

diagnostics