Dev & Coding

A skill that guides Claude to design stable, hard-to-misuse REST/GraphQL endpoints, module boundaries, and type contracts.

A production-engineering playbook for adding structured logs, RED metrics, OpenTelemetry traces, and symptom-based alerts as you build a feature.

A systematic debugging skill that replaces guesswork with reproduce → localize → reduce → root-cause fix → regression test.

A playbook for curating what your coding agent sees, when it sees it, and how it's structured — to keep output quality high.

A shared conventions skill that keeps CHANGELOG.md consistent with Keep a Changelog and SemVer — structure, entry wording, and PR/issue links.

Guides Claude to add or edit MCP server entries in mcptoon's config.json with the correct shape, command rules, and validation loop.

Lets Claude drive the InsForge CLI to provision projects, run migrations, write RLS, manage storage, functions, deployments and payments.

A symptom-driven playbook for diagnosing InsForge backend issues using logs, metrics, DB health, RLS policies, and the advisor.

A prioritized rule set that makes Claude write minimal new code that blends into your existing codebase.

Turns Python CLI scripts into machine-readable NDJSON output behind a --json flag, or scaffolds a full cli_utils package.

A skill that performs multi-file renames and pattern replacements safely using a Grep → Edit → verify workflow.

Generates D1-correct Drizzle ORM schemas, migrations, types, and schema docs for Cloudflare D1.

Scaffolds and deploys a production-ready Cloudflare Worker with Hono routing, Vite dev server, Static Assets, and D1/R2/KV bindings.

A guided workflow for generating, reviewing, applying, and repairing Cloudflare D1 migrations with Drizzle ORM.

Teaches Claude to build semantic search and RAG pipelines on top of the AgentDB vector database.

A reference skill for designing, configuring, and debugging Claude Code hooks — prompt-based and command-based — across all nine hook events.

Teaches Claude the exact syntax and constraints of Apple's `container` CLI for running containers on Apple-silicon Macs without Docker.

A discipline skill that decides whether an addition belongs inside the module, package or service being changed — before any code is written.

Forces ML training failures to be diagnosed by probing for evidence first, then hypothesis, smoke test, control, and only then a root-cause claim.

Forces a probe → hypothesis → smoke → control → claim discipline when an ML training run fails, diverges, OOMs, hangs, or looks slow.