By AdeAnima
Personal coding guardrails: fundamental dev principles (DRY/YAGNI/KISS/SINE), accuracy discipline, milestone review and delegation guidance; a task-workflow skill for repo project management (epics/milestones/tasks in an Obsidian vault) that hands off into superpowers; a warn-only PreToolUse hook that nudges edits into an isolated git worktree; and a blocking PreToolUse hook that routes pre-migration tracker reads into the task-workflow skill.
Engineering judgment for writing and changing code: the four fundamental dev principles (DRY, YAGNI, KISS, SINE), accuracy discipline (never guess APIs), when to run an adversarial review at a milestone, and when to delegate implementation to Codex. Consult this whenever you are about to write, edit, implement, refactor, fix, or review code — including bug fixes, new features, adding a function or class, restructuring, or reviewing a diff or PR. Use it even when the user only says "fix this", "add X", "clean this up", or names a language/framework, and don't wait to be told to apply the principles.
Project task management for any coding repo: epics, milestones, and feature-sized tasks tracked as one markdown note per task in a local Obsidian vault, queried via Obsidian Bases. This is the durable layer ABOVE a single implementation — it owns what work exists and where it stands, then hands each task into the superpowers brainstorm → spec → plan → execute loop and records the spec/plan links. Use it whenever the user is choosing or prioritising what to work on next, grooming or triaging the backlog, adding a feature/epic/milestone, or starting any new feature-sized piece of work. Also applies when a repo already tracks work in an informal tracker (project.md, ROADMAP.md, TODO.md, KNOWN_ISSUES.md, etc.) but has no pm/ vault. Invoke this skill to handle these — do not read tracker files or probe pm/ yourself; the skill owns detection, ranking, and migration. NOT for a one-off typo or single-file bugfix, a two-minute tweak, or ticking an in-flight plan checkbox — those stay in the plan.
Executes bash commands
Hook triggers when Bash tool is used
Modifies files
Hook triggers on file write and edit operations
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Personal Claude Code plugin with coding guardrails. Bundles engineering judgment as an on-demand skill plus a deterministic worktree nudge as a hook — matching each rule to the mechanism that actually enforces it.
| Component | Type | Loads | Purpose |
|---|---|---|---|
coding-principles | skill | on-demand, when doing code work | DRY / YAGNI / KISS / SINE as decision heuristics, accuracy discipline, milestone adversarial review, Codex delegation guidance |
task-workflow | skill | on-demand, when planning/tracking project work | durable PM layer: epics/milestones/feature-tasks as one note per task in an Obsidian vault, queried via Bases, handing off into the superpowers implementation loop |
worktree-gate | PreToolUse hook | every Edit/Write, deterministic | warn-only nudge to isolate edits in a git worktree |
Fires on Edit/Write/MultiEdit/NotebookEdit. Injects a one-time
per-session reminder when you're editing the main working tree of a git repo
instead of an isolated worktree — including on a feature branch, since multiple
agents can share a branch and a shared checkout mixes their work.
It is warn-only: never blocks, always exits 0. Silent inside a linked worktree, in non-git directories, and after the first warning each session.
A durable project-management layer that sits above the per-feature
implementation loop. Work is tracked as one markdown note per feature-sized
task in a local Obsidian vault at pm/, with epics and milestones as their
own notes, all queried via Obsidian Bases (board.base, backlog.base).
The key distinction: a PM task (1–N days, one note, Bases-queryable) is not
a plan step (2–5 min checkbox, ephemeral, owned by superpowers). One PM task
→ one spec file → one plan file. The skill walks a task through
backlog → speccing → planning → in-progress → review → done, handing off to the
superpowers brainstorming → writing-plans → executing-plans loop and
recording the resulting spec_file / plan_file links.
The scripts/tw.sh helper drives it from the command line:
tw.sh init # scaffold pm/ vault + .base views
tw.sh new task "CSV export" --epic epic-export --priority 2
tw.sh next # highest-priority backlog task
tw.sh status T-001 in-progress
tw.sh link T-001 spec docs/superpowers/specs/2026-05-23-csv-design.md
tw.sh check # validate frontmatter for Bases
Obsidian primitives (Bases authoring) come from the
kepano/obsidian-skills plugin;
the implementation loop is superpowers.
This skill owns only the PM layer and the handoff.
The repo is its own marketplace. In Claude Code:
/plugin marketplace add AdeAnima/coding-toolkit
/plugin install coding-toolkit@coding-toolkit
Update later with /plugin marketplace update coding-toolkit then reinstall.
For local development on the plugin itself, load the working copy for one session without installing:
claude --plugin-dir /path/to/coding-toolkit
git and jq on PATH (the hook fails open — stays silent — if jq is
missing).task-workflow uses git and standard POSIX tools (awk/sed); no yq
required. Obsidian (with Bases, v1.9+) to view the vault.MIT — see LICENSE.
npx claudepluginhub adeanima/coding-toolkit --plugin coding-toolkitWeekly meal planning automation: Cookidoo recipe + meal-plan + shopping-list orchestration with German supermarket offer integration (Marktguru)
Personal feature-development workflow chain extending superpowers with dedicated research and multi-subagent review phases
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Develop, test, build, and deploy Godot 4.x games with Claude Code. Includes GdUnit4 testing, web/desktop exports, CI/CD pipelines, and deployment to Vercel/GitHub Pages/itch.io.
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.
Access thousands of AI prompts and skills directly in your AI coding assistant. Search prompts, discover skills, save your own, and improve prompts with AI.