ndisisnd
UserEnd-to-end engineering harness for Claude Code — go from planning to building to reviewing in one workflow. 8 agents with multiple modes. In-built GUI to visualise your PRDs and todos. Ship with AI confidently. MIT.
Categories
Indexed Skills (30)
msg
Root menu for msg skills, plus harness modes. `--init` is the one-time project bootstrap — use it when the user says "initialise project", "bootstrap repo", "set up the framework", "start a new project", or asks to set up project structure in an empty repo. Other modes: `--init-staging` (add a staging branch), `--update` (re-scan a bootstrapped repo), `--gui` (local PRD board), `--doctor` (tally and triage the harness-incident ledger), `--aha` (sweep, triage, and compact the devkit/AHA.md learnings ledger), `--version` (print the installed release in one line), `--help` (guided skill picker).
plan-em
Engineering Manager skill. Reads an approved PRD, runs pre-flight checks against AHA.md, GLOSSARY.md and ARCHITECTURE.md, auto-runs plan-review certification inline (no ask), identifies specialist agents to activate (roster approval — the single human gate), spins them up to write engineering sections directly into the PRD, then synthesises the full output. Sizes itself on the PRD's intake grade: a medium PRD runs one fused plan+build wave behind one certification, a large PRD keeps the two-wave path with a certification before each. Runs in --team mode by default (an Opus orchestrator engineer decomposes each wave into file-disjoint, model-tiered packets fanned out to leaf eng subagents) or --solo (one leaf subagent per roster stack). Refuses without a referenced PRD .md path.
pre-merge
The CI gate. Takes a feature branch from "eng says done" to "PR open against staging with green checks". Runs the project's preflight-resolved pipeline from devkit/policy.json components[]: sync → parallel correctness + security waves → coverage → regression tail → security/migration → PRD-consistency → open PR. Emits a severity-graded verdict JSON. Absorbs the old /review and /test. Activates on /pre-merge after eng --build.
merge
The ship gate — the only skill that merges. `--staging` merges the feature→staging PR on green CI, deploys, verifies, emits a human test script and stamps the staging sign-off on approval. `--production` ships the double-confirmed release to `main` and deploys production. Never self-certifies staging; nothing reaches `main` any other way. Activates on /merge after pre-merge's PR exists.
eng
Platform-agnostic engineering agent with three modes: --plan (propose file changes for human approval AND write the per-feature todo tickets in the same pass), --build (write code from the todo tickets — the single and final build spec), --review (one adversarial whole-change review of the working diff, run by a separate reviewer subagent — spawned by --build automatically, and available standalone on any branch at any time). Use --review when the user says "review this change", "review my diff", "code review this branch", or invokes /eng --review. Invoked by plan-em or directly by the user.
emulate
Runs this repo's app on a local simulator or emulator and opens the window on the user's desktop. Use it when the user says "run the app", "open the simulator", "launch the emulator", "boot the iOS sim", "start Expo", "see it running", "show me the app", or invokes `/emulate`. Takes the branch the user is standing on, clears the stale build processes from the last attempt, and launches. `--ios` / `--adr` / `--expo` pick the lane; with no flag the platform is read from `devkit/PLATFORMS.md`. `--device` pins the simulator or AVD, `--dry-run` shows what would happen and touches nothing. Writes nothing to the repo — no commits, no branches, no PRs.
intake
The planning front-door. Captures feature ideas and bugs as graded rows in the root INTAKE.md ledger. Use it when the user says "log an idea", "capture a bug", "add to the backlog", "note this down", "track this feature", or invokes `/intake`. Owns the requirements interview — fleshes out thin ideas, proactively suggests adjacent ideas, splits compound/hybrid asks and ideas grading C:8 or higher into discrete rows, and grades every idea in a single-turn banded judgment (complexity / token-cost / sequencing). Feeds `plan-pm`, which drafts the PRD. Also owns `--update`: use it when the user says "update that idea", "change the goal on #4", "edit the backlog row about search", or "fix the idea I logged" — it lists the un-shipped rows and edits one in place. And `--delete`: use it when the user says "delete that idea", "remove row #4", "drop that from the backlog", or "I logged that by mistake" — it warns about what the removal breaks, then removes the row on confirmation.
plan-pm
Principal PM skill — the autonomous PRD writer. Consumes a graded, fleshed-out row from the INTAKE.md backlog (idea, goal, type, grade) and drafts the full PRD solo — feature/acceptance table, edge cases, error handling — saved to features/planned/prd-[n]-[feature-slug]/. The requirements interview lives in /intake now, not here. Pauses ONLY for batched open questions the draft couldn't resolve and for breaking/critical touches. Refuses requests that would skip the PRD stage entirely.
plan-review
Staff PM contract certifier. Reads an existing PRD and runs a fixed seven-check certification — each check tied to a named downstream consumer (regression authoring, pre-merge's PRD-consistency gate, the safety pauses, eng --build's row/ticket reads). Product tune (--product) runs checks 1/2/3/6; eng tune (--eng) runs 2/4/5/6/7. Auto-selects the tune type from PRD content (no ask), auto-fixes every Critical and Major with a compact terminal table, asks once about Minors, and pauses only for a product-decision finding. Each auto-fixed Critical/Major writes a category-tagged learning to devkit/AHA.md so the next plan-pm draft self-heals. Applies all fixes directly to the PRD file, and logs its findings to one growing report per PRD at reports/review-prd-[n]-[slug].md.
plan-tune
Staff PM contract certifier. Reads an existing PRD and runs a fixed seven-check certification (G1) — each check tied to a named downstream consumer (regression authoring, pre-merge's PRD-consistency gate, the safety pauses, eng --build's row/ticket reads). Product tune (--product) runs checks 1/2/3/6; eng tune (--eng) runs 2/4/5/6/7. Auto-selects the tune type from PRD content (no ask), auto-fixes every Critical and Major with a compact terminal table, asks once about Minors, and pauses only for a product-decision finding. Each auto-fixed Critical/Major writes a category-tagged learning to devkit/AHA.md so the next plan-pm draft self-heals. Applies all fixes directly to the PRD file. No separate report file.
post-merge
The ship gate. Takes a pre-merge PR from "open against staging" to "live in production". Two modes: `--staging` (verify green CI → merge into staging → deploy → verify per release model — smoke the live target, or submission-accepted for store apps → emit a human test script → stamp staging sign-off on approval) and `--production` (double-confirmed staging→main release PR → merge on green CI + human review → production deploy → verify per release model: smoke the live target, or submission-accepted + monitor-handoff for store apps). The ONLY skill that merges. Never self-certifies staging; nothing reaches `main` any other way. Activates on /post-merge after pre-merge's PR exists.
css
Pragmatic, battle-tested CSS conventions. Use when authoring or reviewing styles — a small set of durable principles (keep specificity low, scope by default, tokenize, respect the user) plus a minimal accessibility/security floor. CSS is forgiving, so this favours judgment and consistency over rigid rules; depth (layout, theming, performance, Tailwind, tooling) lives in refs.
dart
Dart 3.x language standards and code quality conventions. Use when writing or reviewing any Dart code — null safety, patterns, sealed classes, records, class modifiers, naming, immutability, collections, async, and import organisation.
database
Database standards for PostgreSQL persistence and Redis caching. Use when designing schemas, writing migrations, optimizing queries, configuring Redis, or implementing cache invalidation.
flutter
Universal Flutter coding standards. Use when writing or reviewing any Flutter/Dart UI code — widgets, state management, navigation, architecture, performance, or testing.
global
Universal P0 coding standards for any language or stack. Use when writing, reviewing, or refactoring code. Layer- and concern-specific rules live in refs/ and are loaded by cook based on the task.
graphql
GraphQL schema design standards and resolver conventions. Use when writing or reviewing any GraphQL schema, resolver, or operation — naming, nullability, types, input objects, mutations, queries, and operation structure.
macos
macOS app development standards — SwiftUI/AppKit app structure, scenes and windows, menu bar apps, sandboxing and TCC, signing and notarization, distribution, persistence, localization and formatting, and platform conventions (HIG, Liquid Glass, accessibility, undo). Scoped to macOS only — not iOS/iPadOS/watchOS/visionOS.
nextjs
Write and review modern Next.js App Router code with secure RSC boundaries, direct server data access, async route APIs, Server Actions, rendering/cache strategy, and legacy Pages Router awareness.
nodejs
Node.js runtime standards for event loop safety, process lifecycle, streams/backpressure, Buffer memory safety, runtime pinning, package installs, environment loading, and production logging. Use for Node service, CLI, worker, or server runtime implementation and review; co-loads with TypeScript when code is TypeScript.
react
Write correct, idiomatic React with TypeScript. Use when building or reviewing React components, hooks, or JSX — covers hook rules, component structure, TypeScript prop typing, and boundary safety.
supabase
Supabase platform standards — Row-Level Security, publishable/anon and secret/service_role key boundaries, Postgres and Edge functions, Storage, Realtime, and the CLI migration workflow. Use when working with RLS policies, Supabase clients, Edge Functions, or supabase/ migrations. Loads alongside the database (Postgres) domain.
swift
Swift 6.x language standards and code quality conventions. Use when writing or reviewing any Swift code — optionals, error handling, concurrency (actors, Sendable, @MainActor), value types, protocols and generics, memory management, naming, and access control.
typescript
TypeScript 5.x language standards for type safety, narrowing, generics, modules, and async code. Use for TypeScript implementation or review work; load refs only for tooling, testing, or security-specific tasks.
log-it
Check for git commits not yet reflected in CHANGELOG.md, report the count, and offer to write the missing entries. Companion to kermit — run when a session ends before the changelog was updated.
kermit
Format and run git commits using Conventional Commits style with emoji prefix, point-form file bodies, and BREAKING CHANGE footer. Manages CHANGELOG.md (hook-accelerated on Claude Code). Use --init to initialize changelog.
mkpub
Generates and updates a repository's four public-facing docs — README.md, LICENSE.md, SECURITY.md, and llms.txt — from a scan of the repo itself. The README uses a figlet-rendered centered header and badgen badges. Use when the user says "write a README", "generate docs for this repo", "add a license", "add a security policy", "add llms.txt", "/mkpub", or asks to document a repository or skill for release.
cook
React, Next.js, Flutter, Swift, macOS, TypeScript, Node.js, SQL, PostgreSQL, Supabase, GraphQL and CSS coding standards for refactor, audit, bug fix, feature and code review work — this keyword-driven orchestrator receives a code-task summary from the invoking agent, extracts keywords, matches them against the global concern index and the relevant domain indexes, loads only the matched rules, and returns a single compiled standards payload.
nxtup
Opinionated meta-planning for skills, harnesses, and tooling — semver-versioned plans with an approve gate, a --suggest discovery mode, and an orchestrated --build executor. Refuses mobile/web product planning.
wdym
Rewrite, improve, or enhance a prompt. Runs on /wdym, on any request to improve/enhance/rewrite a prompt, and on a UserPromptSubmit <prompt-detect> block. Detects prompt type (code, question, text-gen) and rewrites using 2–3 matched principles; flash mode rewrites immediately, comprehensive mode presents the rewrite and waits for approval. Manage with /wdym --init, --status, --set-mode, --help.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.