flowbit-specs
Spec-driven development rules for Flowbit. Always follow /specs as source of truth.
Best use case
flowbit-specs is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Spec-driven development rules for Flowbit. Always follow /specs as source of truth.
Teams using flowbit-specs should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/flowbit-specs/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How flowbit-specs Compares
| Feature / Agent | flowbit-specs | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Spec-driven development rules for Flowbit. Always follow /specs as source of truth.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Flowbit – Agent Skill ## 1. Source of truth - Always read **all files under `/specs`** before writing or modifying any code. - Treat `/specs` as the **single source of truth**. - If code conflicts with specs, update the code, not the specs (unless explicitly instructed). --- ## 2. Product definition Flowbit is an **Operations OS for SMEs**, not an ERP and not an accounting system. Flowbit manages: - Projects (sales/orders) - Branching workflows using stages (areas) and dependencies - Deadlines per stage and project-level timelines (planned vs actual) - Unit quantity tracking per stage (qty_required, qty_done) - Visual finance using **income/cost events** (no formal accounting) - Products with BOM (materials per unit) and **project-level editable snapshots** Out of scope: - Formal accounting (PUC, taxes, invoices) - Payroll - Complex inventory or reservations - Automated notifications - Over-engineered abstractions --- ## 3. Architecture & stack ### Backend - FastAPI - SQLAlchemy - Pydantic - PostgreSQL - Multi-tenant architecture: **every entity is scoped by `company_id`** ### Frontend - React + TypeScript - Vite - Tailwind CSS - **shadcn/ui** components (Radix-based) - Icons from `lucide-react` --- ## 4. UI rules (very important) - Build a **real SaaS dashboard**, not a demo or toy UI. - Use sidebar + top header layout. - Prefer `Card`, `Table`, `Tabs`, `Badge`, `Button`, `Dialog` from shadcn/ui. - Avoid raw `<div>`-only layouts. - Avoid inline styles. - Pages must look production-ready and consistent. --- ## 5. Roles & permissions Roles: - `SUPER_ADMIN`: platform owner (can create companies). - `COMPANY_ADMIN` (Boss): full access inside a company, including finance. - `STAGE_WORKER`: access only to assigned stages. Rules: - Finance data is visible **only** to `COMPANY_ADMIN`. - `STAGE_WORKER` users must never see: - prices - margins - income or cost events - Stage access is mandatory to view or edit a `ProjectStage`. --- ## 6. Workflow rules (branching) - Projects select which stages apply. - Stages have dependencies (`edges`) that define prerequisites. - Stage statuses: - `BLOCKED` - `READY` - `IN_PROGRESS` - `DONE` - `SKIPPED` - A stage becomes `READY` only when **all selected prerequisites are `DONE` or `SKIPPED`**. - When a stage becomes `DONE`, dependent stages must be re-evaluated. --- ## 7. Deadlines & timeline Each `ProjectStage` tracks: - `planned_due_date` - `actual_ready_at` - `actual_started_at` - `actual_done_at` Timeline rules: - Show planned vs actual per stage. - Delays must be computable from stored timestamps. - Timers are optional but supported. --- ## 8. BOM & materials - Product BOM defines materials per unit. - On project creation, generate **project material requirement snapshots**: - quantities per unit - total quantities - Project material requirements are **editable**. - Editing project requirements must **not** modify the product BOM. --- ## 9. Finance (visual only) - Financial events are simple records: - type: `INCOME` or `COST` - amount - date - category - No accounting logic, no double entry, no taxes. - Metrics: - total income per project - total cost per project - margin = income − cost --- ## 10. Development philosophy - MVP first. Do not build future features prematurely. - Prefer clarity over cleverness. - Avoid over-abstraction. - Code should look **production-ready**, not tutorial code. --- ## 11. When in doubt If a requirement is unclear: 1. Check `/specs`. 2. If still unclear, ask before implementing.
Related Skills
dev-swarm-tech-specs
Define technical specifications including tech stack, security, theme standards (from UX mockup), coding standards, and testing standards. Use when user asks to define tech specs, choose tech stack, or start Stage 7 after architecture.
agent-file-specs
Contains the complete specifications for AI coding assistant customization files including agents, skills, prompts, and instructions. Works with GitHub Copilot, Claude Code, Codex, OpenCode, and other providers. Use this skill when you need to reference the correct file format, required fields, supported attributes, file locations, or VS Code settings for any customization file. Follows the Agent Skills open standard (agentskills.io).
bgo
Automates the complete Blender build-go workflow, from building and packaging your extension/add-on to removing old versions, installing, enabling, and launching Blender for quick testing and iteration.
poetry-rye-dependency-management
Specifies Poetry or Rye for dependency management in Python projects.
podcastfy-clawdbot
Generate an AI podcast (MP3) from one or more URLs using the open-source Podcastfy project. Use when the user says “make a podcast from this URL/article/video/PDF”, “turn this webpage into a podcast”, or wants an MP3 conversation-style summary from links. Uses Gemini for transcript generation via GEMINI_API_KEY and Edge TTS for free voice.
plugin-patterns
Canvas plugin architecture patterns, best practices, and implementation templates
playwright-skill
Complete browser automation with Playwright. Auto-detects dev servers, writes clean test scripts to /tmp. Test pages, fill forms, take screenshots, check responsive design, validate UX, test login flows, check links, automate any browser task. Use when user wants to test websites, automate browser interactions, validate web functionality, or perform any browser-based testing.
playwright-core
Battle-tested Playwright patterns for E2E, API, component, visual, accessibility, and security testing. Covers locators, assertions, fixtures, network mocking, auth flows, debugging, and framework recipes for React, Next.js, Vue, and Angular. TypeScript and JavaScript.
playwright-app-testing
Test the Expensify App using Playwright browser automation. Use when user requests browser testing, after making frontend changes, or when debugging UI issues
platform-engineer
**Master Skill**: Unified Platform, SRE & Release Engineering. Covers OpenShift 4.20+, GitOps (ArgoCD/Tekton), Container Hardening, Service Mesh, Feature Flags, Progressive Rollouts, Observability (LGTM Stack), Chaos Engineering, and Disaster Recovery.
platform-detection
Detect project type and recommend deployment platform. Use when deploying projects, choosing hosting platforms, analyzing project structure, or when user mentions deployment, platform selection, MCP servers, APIs, frontend apps, static sites, FastMCP Cloud, DigitalOcean, Vercel, Hostinger, Netlify, or Cloudflare.
platform-backend
Server-side architecture and security patterns. Extends core-coding-standards with API, error handling, and security rules. Use when building APIs or server logic.