superpowers
claudecodexCore skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
Describe what you want in plain language. onestop plans the phases and dispatches the right specialist for each - requirements, design, code, tests, web and app automation, review, ship - stopping at every gate for your approval. A pipeline engine enforces the gates, retry budgets and safety rules in code; technology and version choices and everything that reaches git stay yours.
Plugin onestop from marketplace onestop in karthikeyabommireddy/oneStop. Commands read from the repo's manifest — 1 month ago.
This plugin runs code on your machine — its hooks fire automatically on events like tool use and it starts local MCP servers. The exact commands are listed below; read them before installing.
orchestrateThe onestop hub that runs a request phase by phase through the engine. Use only when /onestop or /onestop-resume invokes it.
parallel-executionThe wave scheduler. Turns a task DAG into waves of independent agents that run at the same time, joins each wave, and continues around failures. Shared engine used by the discovery, implement, test, automation and review phases. Loaded by the orchestrate skill; not usually invoked directly.
phase-automationBuild end-to-end automation for the described journey across web and native app targets. Binds the existing or default framework per target, delegates to the web and app automation agents, and wires artifacts and CI. Loaded by the orchestrate skill; not usually invoked directly.
phase-complianceCheck a change against the regulatory regime its data and domain actually place it under - HIPAA, GDPR, PCI-DSS, SOC 2 - and report control findings with severity. Loaded by the orchestrate skill when a regulated data path is detected.
phase-contextEstablish what the repository is before any phase reasons about it - the detected stack, the project's own build, test and lint commands, the conventions, and the code knowledge graph - recorded once as stack facts every later specialist reads. Loaded by the orchestrate skill; not usually invoked di
phase-designProduce the design artifacts a change needs and no more - interface contract, low-level design, and an ADR for decisions that are expensive to reverse. Scales from a single contract file to a full HLD. Loaded by the orchestrate skill; not usually invoked directly.
phase-discoverySearch the repository to establish what already exists before any planning or questioning happens. Runs one scout pass per unit of work, aggregates the findings, and hands the orchestrator a resolved decision set plus the short list of choices that genuinely need the user. Loaded by the orchestrate
phase-gateWhy onestop gates every phase boundary, and the edge cases - gate zero, skips, adjustments, blocked phases, user-only decisions, gate modes. The format itself is orchestrate Step 5; the engine records and enforces every gate. Loaded by the orchestrate skill; not usually invoked directly.
phase-implementExecute the approved task list test-first, one vertical slice at a time, following the conventions discovery recorded. Drives the red-green-refactor loop, repairs build breaks in place, and keeps the suite green between slices. Loaded by the orchestrate skill; not usually invoked directly.
phase-measureProduce a reproducible performance baseline before any optimization, so improvement can be proven rather than assumed. Loaded by the orchestrate skill for perf intent.
phase-planTurn the discovery picture into an ordered task list of thin vertical slices, each with its own acceptance criteria and test strategy. Produces the artifact presented at the plan gate. Loaded by the orchestrate skill; not usually invoked directly.
phase-qa-planDraft the manual testing plan for the change as a document plus an importable CSV, covering what a human tester does on top of the automated suites. Runs after automation, before review. Loaded by the orchestrate skill; not usually invoked directly.
phase-reproduceTurn a reported defect into a failing regression test that fails for the right reason, before any fix is attempted. Loaded by the orchestrate skill for defect intent; not usually invoked directly.
phase-requirementsTurn a thin request, ticket, or spec document into testable acceptance criteria before planning begins. Runs only when the input lacks criteria a plan can be built on. Loaded by the orchestrate skill.
phase-researchFind proven prior art before writing net-new code - vendor documentation, package registries, and reference implementations - and report what should be adopted rather than built. Loaded by the orchestrate skill; not usually invoked directly.
phase-reviewReview the change with every reviewer the diff surface warrants, running them in parallel, and resolve blocking findings before ship. Binds language, security, database, accessibility and domain reviewers automatically from the diff. Loaded by the orchestrate skill; not usually invoked directly.
phase-scaffoldStand up the first end-to-end vertical slice of a new project from the ecosystem's official generator, at the language, framework and versions the user chose, so every later slice has a working skeleton to extend. Loaded by the orchestrate skill for mvp intent.
phase-shipClose the loop after review - conventional commits scoped to logical changes, a pull request with the repo template, and documentation sync. Owns the ship gate, the last stop before anything leaves the machine. Loaded by the orchestrate skill; not usually invoked directly.
phase-testComplete unit and integration coverage for the change, verify the coverage threshold, and prove the tests are correct rather than merely present. Loaded by the orchestrate skill; not usually invoked directly.
phase-ui-designEstablish or extend the visual design system before UI implementation - domain-matched palette with verified contrast, type scale, spacing, elevation and motion, emitted as tokens. Runs automatically when the change touches user interface. Loaded by the orchestrate skill; not usually invoked directl
phase-upgrade-planPlan a dependency, runtime or framework upgrade - current and available versions from the official registry, the breaking changes between them, the call sites they hit, and the user's choice of target version as the gate that authorises implementation. Loaded by the orchestrate skill for upgrade int
phase-verify-greenProve the existing suite is green and actually covers the code about to be restructured, before a refactor begins. Adds characterization tests where coverage is missing. Loaded by the orchestrate skill for refactor and upgrade intent.
/onestop:onestop-helpWhat onestop can do, with examples. Starts nothing.
/onestop:onestop-resumeResume the current onestop run, optionally with a correction.
/onestop:onestop-statusShow where the current onestop run stands.
/onestop:onestop-undoUndo the onestop run's last change, or the whole run, after a preview.
/onestop:onestopRun a request through onestop - phase by phase, with your approval at each gate.
a11y-agentAccessibility specialist: WCAG 2.2 AA design and review of web and native UI.
app-automation-agentNative and cross-platform app end-to-end automation - Detox, Espresso, XCUITest, FlaUI, Flutter.
architectDesigns module boundaries, data model, contracts and ADRs; scaffolds new projects.
ba-analystTurns a request into requirements, stories with testable criteria, and a traceability matrix.
build-resolverGets a failing build or type check green with the smallest correct change.
code-explorerTraces how existing code actually works, with file-level evidence. Read-only.
code-reviewerReviews a change for correctness, contracts, errors and fit, in every language of the diff. Read-only.
contract-agentWrites the interface contract both sides of a boundary build against.
data-reviewerReviews migrations, schema and queries for deployment safety and performance. Read-only.
devops-agentFixes and changes CI and build infrastructure in the repository's own system.
discovery-scoutFinds what already exists in the repository before anything is asked. Read-only.
docs-agentUpdates documentation from the source of truth, only where the change invalidated it.
implementerWrites the production code that makes a failing test pass, inside its write surface.
merge-coordinatorJoins a parallel wave: verifies the combined state and catches semantic conflicts.
option-brokerDecides which open choices must go to the user, and frames them with a recommendation.
performance-agentMeasures, profiles and fixes real performance problems against a baseline.
plannerTurns discovery and design into thin vertical slices with acceptance criteria.
qa-plannerWrites the manual test plan and importable test cases.
refactor-agentRestructures code without changing behaviour, suite green at every step.
researcherFinds versions, documentation and breaking changes from official sources, verified today.
security-reviewerReviews a change for exploitable weaknesses, and checks compliance controls.
stack-adapterConfirms the stack, commands and conventions every later specialist relies on.
test-authorWrites tests first and runs the red-green-refactor loop in the stack's idiom.
ui-designerBinds a domain palette, visual style and design tokens, with measured contrast.
validatorRenders the final pass or fail verdict on a completed change.
web-automation-agentBrowser end-to-end automation from a described user flow, using the live page.
work-partitionerSplits an approved plan into tasks with write surfaces and parallel waves.
engineruns: node ${CLAUDE_PLUGIN_ROOT}/engine/server.mjs
chrome-devtoolsruns: npx -y [email protected] --isolated
playwrightruns: npx -y @playwright/[email protected] --isolated
context7runs: npx -y @upstash/[email protected]
node "${CLAUDE_PLUGIN_ROOT}/engine/hooks.mjs" pre-tool
node "${CLAUDE_PLUGIN_ROOT}/engine/hooks.mjs" post-tool
node "${CLAUDE_PLUGIN_ROOT}/engine/hooks.mjs" subagent-stop
node "${CLAUDE_PLUGIN_ROOT}/engine/hooks.mjs" stop
node "${CLAUDE_PLUGIN_ROOT}/engine/hooks.mjs" session-start
node "${CLAUDE_PLUGIN_ROOT}/engine/hooks.mjs" prompt-submit
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
Backend API design, GraphQL architecture, workflow orchestration with Temporal, and test-driven backend development
9 skills · 1 command · 8 agents
AI-supervised issue tracker for coding workflows
1 skill · 1 agent · hooks
Use Superset to spawn agents, create workspaces, manage tasks, and schedule automations from Claude Code.
14 skills · 1 command
MiniMax AI skills library for frontend, fullstack, Android, iOS, shader, GIF sticker, document, presentation, spreadsheet, and multimodal media workflows
18 skills
Worktrunk is a CLI for Git worktree management, designed for parallel AI agent workflows. This plugin provides configuration guidance (LLM commit messages, project hooks, worktree paths), automatic activity tracking (🤖/💬 indicators in `wt list` showing active Claude sessions), and a `/wt-diff` pane showing each worktree's changes.
2 skills · hooks
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.