A highly opinionated plugin marketplace for augmenting your Claude Code. Battle-tested skills and tools that actually work.
Agent
Use this agent when you need architectural guidance or review. Recommended PROACTIVELY after adding features that span 3+ modules, refactoring cross-module boundaries, or making structural design decisions. Not for naming, formatting, or single-file style questions — those are code quality, not architecture.
Agent
Use this agent when you need code quality review — readability, correctness, dead code, naming, and consistency. Use PROACTIVELY after implementing features, fixing bugs, or refactoring. Focuses on changed code and its immediate context. Not for security (use security-auditor), performance (use performance-auditor)…
Agent
Use this agent when you need to find performance bottlenecks — algorithmic complexity, inefficient data access, or resource-usage problems. Recommended PROACTIVELY after adding loops over large or unbounded data, queries inside loops, or hot-path changes. Not for code quality (use code-auditor), security (use…
Agent
Use this agent when you need a security review — finding exploitable vulnerabilities in authentication, untrusted-input handling, secrets, or dependencies. Recommended PROACTIVELY after changes to auth/authz, API endpoints, input parsing, file uploads or path handling, or cryptography. Not for general code quality…
Agent
Use this agent when you need tests for implemented features or when test coverage analysis is needed. Recommended PROACTIVELY after implementing features, fixing bugs, or adding new modules. Focuses on critical business logic testing, not coverage metrics. Not for code quality review (use code-auditor) or performance…
Agent
Use this agent when you need to verify existing memory entries against codebase ground truth — checking for stale paths, outdated versions, contradicted facts, and relative dates — and to prune the corpus by flagging superseded, redundant, and low-value entries. Recommended PROACTIVELY after large refactors or version…
Agent
Use this agent when the user wants to mine recent conversation sessions for uncaptured knowledge — corrections, architectural decisions, recurring patterns, and behavioral preferences. Not for persisting memories — use extract-learnings for that.
Agent
Use this agent when a skill needs structural linting after creation or improvement. Trigger PROACTIVELY after create-skill or improve-skill completes. Audits SKILL.md against the 7-dimension quality standard; auto-applies critical and major fixes. Not for agent files — those have a different contract.