wai-suite

Community

Sustainable software product development with Claude Code: 13 skills for the engineering lifecycle — plan, build, test, review, audit — built to keep architecture maintainable and clean and security first-class while agents write the code. The working premise, drawn from field runs: even frontier…

Plugin · by woldinius · runs in Claude Code, Cowork, claude.ai · 13 skills, 1 agent · listed 2026-08-12

Claude Code

claude plugin install wai-suite

A community plugin: Anthropic has not reviewed it. Read the source before you install it.

Source

Skills (13)

Skill wai

Front door and router for the wAI skill suite. Use it to orchestrate — when you are unsure which skill to start with, want a lifecycle overview, or a recommendation on what to run next: "which skill…

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai
```
Skill wai-architecture-audit

Periodic, whole-codebase structural-health audit: architectural drift, coupling, god-services, dead code — and the subtle part, semantic redundancy, inconsistency and cross-surface dead-ends…

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-architecture-audit
```
Skill wai-cicd

One-time setup of GitHub-native CI/CD for a backend+web repo (Actions, GHCR, Docker Compose over SSH to your own Linux server, any provider) — and owner of the merge gate that makes "green checks"…

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-cicd
```
Skill wai-implementation

Default skill for implementing changes — the senior fullstack doer. Plans first (risk and blast radius), builds on the requirement's `agent/**` branch, self-reviews, and opens the PR.

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-implementation
```
Skill wai-init

Bootstrap and re-runnable updater for the wAI suite: deep-scans a repo, then writes its surface-scoped quality catalog (`docs/architecture/quality-attributes.md`) and testing strategy, sized to the…

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-init
```
Skill wai-learning-gap

Personal learning mode (cloze coding + Leitner), tech-stack agnostic — opt-in per developer, never repo-wide. Active ONLY for a human who has a personal ledger under `~/.claude/learning/`; with no…

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-learning-gap
```
Skill wai-mobile-release

One-time setup of build, code signing and store delivery for a native iOS or Android repo (TestFlight/App Store Connect, or Play Console tracks) — and owner of the mobile merge gate.

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-mobile-release
```
Skill wai-pr-review

Senior architecture/DevOps review of a pull request or diff against the quality catalog and the originating spec, ordered by severity — and owner of the gated auto-merge.

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-pr-review
```
Skill wai-requirements-planning

Plans and shapes a new requirement before implementation — architecture, DevOps and the cross-cutting concerns (cost, GDPR, API compatibility, resilience, token economy) surfaced early, with a…

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-requirements-planning
```
Skill wai-retro

Periodic retrospective of the suite's own work, built from artifacts at a threshold — never from recall. Reads the gate ledger's report extract, the run log and git history; writes a dated, narrated…

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-retro
```
Skill wai-security-audit

Periodic, whole-codebase security audit — an adversarial "how would an attacker break this" sweep: authN/authZ and IDOR, secrets, injection (incl.

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-security-audit
```
Skill wai-team

Mandated lifecycle orchestrator: works ONE or MORE GitHub issues through the full lifecycle — per issue one run of plan → implement → test → review on its own `agent/**` branch and PR, integrated…

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-team
```
Skill wai-testing

Writes deterministic tests and owns the testing strategy. Mandatory targets are security, billing, GDPR and idempotency; tests never call a real model or real billing.

Any agent
```
npx skills add woldinius/wai-skill-suite --skill wai-testing
```

Agents (1)

Agent wai-reviewer

The suite's fresh-context reviewer — reviews ONE pull request (or a local diff) with the wai-pr-review skill preloaded, posts the review, runs the merge gate and adds its result, and merges nothing.

From Anthropic's public plugin directory, read on 2026-10-15. Find a skill by task.