Compares implementation (routes, types, status codes) to contract or spec; reports drift and suggests fixes or spec updates. Use when the user mentions "contract," "spec," "OpenAPI," or "does this match the API.".
Evaluates architecture options using explicit tradeoffs (scalability, coupling, operability, delivery risk) and recommends a decision with rationale. Use when choosing between system designs, technologies, or topology options.
Plans architecture-level evolution from current state to target state with compatibility, migration phases, and risk controls. Use when refactoring major subsystems, splitting services, replacing infrastructure, or addressing architecture drift.
Selects architecture patterns only when justified by concrete failure modes or scaling pressure; rejects pattern cargo-culting. Use when deciding service boundaries, layering, repository/service patterns, CQRS, events, or modularization strategy.
Adds mandatory inline comments and docstrings explaining why, invariants, and non-obvious logic when writing or editing code. Use when the user asks to "add comments," "document this code," "explain this function," or when shipping new or materially changed logic per ai-guardrails.
Adds specific raises, narrow catches, exception chaining, and domain error types when writing or editing code. Use when the user asks to "add error handling," "raise specific exceptions," "custom exceptions," or when shipping new or materially changed logic with raise/except paths per ai-guardrails.
Adds structured code-path logging (entry/exit, errors, key decisions); applies log levels, follows project patterns, and redacts sensitive data. Use when the user asks to "add logging," "log this," or make a code path debuggable through logs.
★not rated 5 2mo agoA60 tokens
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: