Goldziher

Goldziher/ai-rulez

Agent Claude Code

Use when writing tests — follows TDD red-green-refactor cycle.

Goldziher/ai-rulez

Skill Claude Code

Standards-compliant lifecycle tool for agent knowledge and capabilities: author in .ai-rulez/, generate for 52 harnesses (Claude Code, Cursor, Codex, Copilot, Gemini CLI, OpenCode, Devin, and more), bundle (OKF, Agent Plugins, ARD), validate, govern (lock, approvals, signing, SBOM) and publish. Use when configuring…

Goldziher/ai-rulez

Skill Claude CodeCodex

Concrete code-quality thresholds and anti-patterns: 120-char lines, max 20 cyclomatic complexity, max 4 nesting levels, max 50 lines per function, no magic numbers, no global state, composition over inheritance, no bare except, rule-of-three before extracting an abstraction, delete dead code instead of commenting it…

Goldziher/ai-rulez

Skill Claude CodeCodex

Language-idiomatic error handling: wrap errors with the failed operation's context, never swallow silently, Result in Rust, fmt.Errorf("doing X: %w", err) in Go, typed exceptions in Python/Java, fail fast on unrecoverable errors. Load when writing or reviewing an error path, a catch/except block, a custom error type…

Goldziher/ai-rulez

Skill Claude CodeCodex

Container build and runtime standards: multi-stage builds into a minimal runtime (Alpine/Distroless/Scratch), manifest-before-source layer ordering for cache hits, non-root user, vulnerability scan failing on HIGH/CRITICAL, no secrets in the image, tini/dumb-init as PID 1, HEALTHCHECK, and semver+SHA tags that are…

Goldziher/ai-rulez

Skill Claude CodeCodex

Observability standards: JSON structured logging with key=value fields and never string interpolation, the ERROR/WARN/INFO/DEBUG/TRACE level taxonomy, a /health endpoint with per-component status wired to orchestrator probes, counter/gauge/histogram metric choice with no high-cardinality labels, and correlation-ID…

Goldziher/ai-rulez

Skill Claude CodeCodex

Test-driven development discipline: RED (failing test) → GREEN (minimal code) → REFACTOR, delete-and-restart if production code was written first, and the test-type taxonomy (integration for API surfaces, unit for business logic, property for edge-case-heavy code). Load before starting a feature or bug fix, when…

Goldziher/ai-rulez

Skill Claude CodeCodex

How to write a good test: behaviour-describing names (shouldreturnerrorwheninputisempty, givenwhenthen), exact-value assertions over truthiness, snapshot and property-based testing, order-independent and idempotent tests with no shared mutable state, and the anti-patterns (testing mocks, test-only production methods…

Goldziher/ai-rulez

Skill Claude CodeCodex

Build in verified increments rather than one speculative block: smallest viable change first, verify, then extend; prototype the risky part before committing to a design; batch edits to one file into a single operation. Load when planning a large or multi-file change, when uncertain which of several approaches will…

Goldziher/ai-rulez

Skill Claude CodeCodex

Use task (go-task) instead of raw build/test/lint commands in any repository whose root contains a Taskfile.yaml or Taskfile.yml, with the standard target names setup, build, test, lint, format, bench. Load on seeing a Taskfile.yaml/Taskfile.yml in the repository root, or before running a build, test, lint or format…

Goldziher/ai-rulez

Command Claude Code

Author, generate, validate and govern AI agent rules, skills, plugins and MCP configuration across 52 coding assistants. Review with verified model evidence, evaluate skills, lock and sign dependencies, and publish bundles. CLI, MCP server and Go API.

Goldziher/ai-rulez

Command Claude Code

Opens a pull request for the current branch with the ticket key in the title.

Goldziher/ai-rulez

Skill Claude Code

Drafts release notes from merged pull requests. Use when a release is being prepared.

Goldziher/ai-rulez

Skill Claude Code

Bumps one dependency to a newer version. Use when the user names a package to upgrade.

Goldziher/ai-rulez

Skill Claude Code

Sets up the JavaScript toolchain on a fresh checkout. Use when onboarding a new machine.

Goldziher/ai-rulez

Skill Claude Code

Regenerates the dependency lockfile. Use when the manifest changed and the lockfile is out of date.

Goldziher/ai-rulez

Skill Claude Code

Checks that log rotation is configured. Use when a service disk fills up.

Goldziher/ai-rulez

Skill Claude Code

Names a new git branch. Use when the user starts work on a ticket.

Goldziher/ai-rulez

Skill Claude Code

States the secrets policy for contributors. Use when asked how credentials are handled.

Goldziher/ai-rulez

Skill Claude Code

Commits the finished changes as one commit. Use when the user says the work is ready to commit.

Goldziher/ai-rulez

Skill Claude Code

Creates the git branch for a ticket. Use when the user starts a ticket and gives its key and a short slug.

Goldziher/ai-rulez

Skill Claude Code

Writes a unit test for one Python function. Use when the user asks for tests of existing code; not for integration tests.

Goldziher/ai-rulez

Agent Claude Code

Checks React components for missing accessibility attributes. Use when a component is added or changed.

Goldziher/ai-rulez

Command Claude Code

Scaffolds a new React component next to an existing one.