plaesy/spec-kit

Angular coding standards and best practices. Use when building or reviewing Angular components, services, templates, or styles.

Architecture design principles: API, caching, DevOps, CLI, concurrency, database, distributed systems, domain-driven design, observability, resilience, event-driven, monorepo, and general software design. Use when designing or reviewing a system; load only the file that matches the decision.

Search arXiv papers by keyword, author, category, or ID; read abstracts and full papers; generate BibTeX entries; enrich with Semantic Scholar citation data, recommendations, and author profiles. Use when the user asks to find, read, cite, or analyze arXiv papers.

Assessment workflows per dimension: business, design, financial, legal, management, marketing, operations, product, technical. Use when the user asks to assess or audit one dimension; load the file for that dimension.

Track buyer outreach, interviews, validate demand.

BPOM/FTC health claims compliance checker with DB.

Context7 documentation lookup: when it is mandatory, the exact tool sequence, the citation form, and the fallback. Use when the user asks about a library, framework, SDK, or API, or wants current docs or examples.

Intelligent state detection and continuation orchestrator Use when the user types /continue.

C# coding guidelines for building .NET applications. Use when writing, reviewing, or refactoring C# (.cs) code.

Dart and Flutter coding standards from the official recommendations. Use when writing or reviewing Dart or Flutter code.

UI and frontend design: aesthetic direction and typography for new UI, upgrading existing UI, accessibility (WCAG, screen readers, contrast), UI/UX principles, and critiquing a rendered page. Use when the user asks for UI or frontend design, accessibility fixes, or a visual review; load the file that matches the task.

Generate grounded, cited project documentation (overview, architecture, components, reference, how-to) from source code and specifications. Use when the user asks to document a project, write docs or an API reference, or prepare a project for handoff

Auto-calculate unit economics, break-even, verdict.

Create or edit Excel (.xlsx) workbooks with openpyxl or XlsxWriter: sheet structure, formulas, formatting, and reports. Use when the user asks for a spreadsheet or .xlsx output.

Find or test a business idea before launch. Suggests candidate ideas from researched market gaps when the user has none, then walks the chosen idea through a board review, competitor and buyer research, pricing and offer, unit economics, marketing, brand, operations and launch, and compiles one business plan whose verdict is computed from the numbers. Use when the user says "find me a business idea", "suggest ideas", "test my business idea", "is this worth building", "write my business plan", "pressure-test this idea", or asks whether a business is profitable before they start.

Git workflows, branching strategies, commit discipline, and version control practice. Use when creating branches, commits, merges, tags, or pull requests.

Idiomatic Go practices and community standards for services, CLIs, and libraries. Use when writing, reviewing, or refactoring Go (.go) code or a Go service.

Generate divergent business ideas with assumptions, tests.

Java best practices, bug patterns, code smells, and build verification. Use when writing or reviewing Java code. Spring Boot rules are in ../spring-boot/SKILL.md.

Kubernetes deployment best practices for container workloads: Pods, Deployments, Services, Ingress, ConfigMaps, Secrets, probes, resource limits, scaling, and security contexts. Use when writing or reviewing Kubernetes manifests or deploying containers to Kubernetes.

Markdown (.md) documents: structure, formatting, and accessibility. Use when creating or editing a .md file, README, or markdown page.

NestJS standards for scalable Node.js server-side applications: modules, providers, controllers, and testing. Use when building or reviewing NestJS services.

Next.js standards for the App Router, route handlers, and server and client components. Use when building or reviewing Next.js applications.

Performance work: optimization for frontend, backend, and database, and the baseline measurement framework that proves a change helped. Use when the user asks to speed up, profile, or benchmark code; load references/baseline.md before optimizing and references/optimization.md when choosing changes.

Run full pipeline from niche to make money verdict.

Write the complete implementation plan for a topic: how each requirement in its spec will be met and verified, with ordered steps, dependencies, risks, a rollback path, and a go/no-go gate. Use when the user asks for a plan, an implementation approach, or a rollout, after the spec exists and before any code changes.

Compile founder stage files into plan.md with verdict.

Generate and edit PowerPoint (.pptx) files programmatically with python-pptx. Use when the user asks for slides, a presentation, or .pptx output.

Python standards: PEP 8 style, typing, packaging, testing, async, and common anti-patterns. Use when writing, reviewing, or refactoring Python code.

React standards and best practices. Use when building or reviewing React components, hooks, or state.

React Native standards for mobile apps, with Spec-Kit constitutional compliance. Use when building or reviewing React Native code.

Research a question with cited sources before a decision: find the sources, open them, record each claim with its URL and access date, and label anything unverified. Use when the user asks to research, compare options, check facts, find current prices or versions, or when a spec, plan, or decision needs evidence.

Ruby on Rails coding conventions for models, controllers, routes, and migrations. Use when writing or reviewing Rails code.

Idiomatic Rust practices and community standards. Use when writing, reviewing, or refactoring Rust (.rs) code.

Save session state and update memory checkpoints Use when the user types /save.

Secure coding and security audit: OWASP Top 10 (2021) rules, deprecated-library prohibition, dependency verification, WCAG 2.2 AA checks, and the audit report format. Use when writing security-sensitive code or auditing for security or accessibility issues.

Create a new Agent Skill (a folder with a SKILL.md) that follows current best practice: correct frontmatter, a description that makes the skill easy to find, a concise body, and optional reference, script and asset files. Use when the user asks to create, write, or scaffold a skill, or to turn a repeated workflow into one.

Write the complete spec for a topic: requirements R1..Rn with Given/When/Then acceptance criteria, non-goals, constraints, and open questions. Adds design.md and one self-contained HTML mock per page for any work with a user interface, and api.md for any work with an interface other agents or services call. Use when the user asks to write a spec, define requirements, or describe what a feature must do, before any plan or code.

Write the business spec for a topic: market, customers, value proposition, business model, revenue, and success measures, with requirements in R1..Rn and Given/When/Then criteria. Use for a new business, a new market, or a new business model, before any plan or pitch. Use when the user asks to write a business spec, business requirements, or a business case.

Write the financial spec for a topic: budget, cost structure, pricing, cash flow, and the financial limits the work must respect, with requirements in R1..Rn. Use when the user asks for a budget spec, a cost model, pricing requirements, or financial limits.

Write the legal and compliance spec for a topic: applicable laws and regulations, data and privacy obligations, contract terms, and required approvals, with requirements in R1..Rn. Use when the user asks for a legal spec, compliance requirements, contract requirements, or a privacy review.

Write the management spec for a topic: owners, roles, decisions, schedule, and communication, with requirements in R1..Rn. Use when the user asks for a management spec, a RACI, a project charter, or a schedule and ownership plan.

Write the marketing spec for a topic: audience, message, channels, campaign goals, and measures, with requirements in R1..Rn and Given/When/Then criteria. Use when the user asks for a marketing spec, a campaign brief, or go-to-market requirements.

Write the operations spec for a topic: how the work is run, deployed, supported, and monitored, with requirements in R1..Rn. Use when the user asks for an operations spec, a runbook requirement, an SLA, or support requirements.

Write the product spec for a topic: users, problems, features, priorities, and what is out of scope, with requirements in R1..Rn. Use when the user asks for a product spec, a PRD, a feature list, or product priorities.

Write the quality spec for a topic: what must be tested, the acceptance tests for each requirement, environments, and exit criteria, with requirements in R1..Rn. Use when the user asks for a test spec, QA requirements, a test plan, or acceptance criteria for a release.

Spring Boot guidelines: dependency injection, configuration, code organization, service layer, logging, and input handling. Use when building or reviewing Spring Boot services.

SQL guidelines for writing, fixing, and optimizing queries, schemas, and stored procedures. Use when the user asks to write, review, fix, or speed up a SQL query or statement.

Static website standards (plain HTML/CSS/JS or a static-site generator): performance, SEO, and deployment. Use when building or reviewing a static site.

Create, start, complete, block, unblock, and check Plaesy tasks. Each task is a folder under .plaesy/tasks/<status>/ with a task.md and a contracts.json, and belongs to a topic whose spec and plan live under .plaesy/spec/ and .plaesy/plan/. Use when the user asks to create a task, start or finish work on one, block or unblock it, check whether it is valid, or see what is in progress.

Terraform conventions for file layout, naming, variables, outputs, and providers. Use when writing or reviewing .tf files or Terraform modules.

Testing discipline: test-driven development (red-green-refactor, unit tests, a commit per step) and the testing strategy (what to test, golden master and characterization tests, coverage). Use when writing tests, adding a regression test, planning test coverage, or validating a change.

Reduce token use in agent sessions without losing information: short answers that keep code and paths exact, fewer repeated file reads, and compact command output through scripts/compact_output.py. Use when a session is long or costly, when a command prints a large log or test run, or when the user asks to cut tokens.

Bootstrap for every session in a Plaesy project. Load at the start of any task to find and apply the skill that matches it before acting.

Vue standards: Composition API, Single-File Components, reactivity, props and events, state, testing, and performance. Use when building or reviewing Vue components.

Generate and edit Word (.docx) documents with python-docx: reports, letters, and formatted documents. Use when the user asks for a Word document or .docx file, not for markdown.

Automate YouTube video creation from topic to publish.