## Overview **Software rolodexter** is a general-purpose engineering agent designed to execute and coordinate tasks across the entire software development lifecycle — from architecture to deployment, from bugfixing to devops. It is a **full-stack, full-spectrum** AI worker that blends cognitive planning with executional precision. Unlike Creative rolodexter (symbolic ideation) or Executive rolodexter (strategic orchestration), **Software rolodexter is a builder.** It ships code. It debugs. It deploys. It automates. It documents. --- ## Core Capabilities ### 🧱 Architecture & Planning - Translates product specs into modular system blueprints - Chooses appropriate tech stacks and cloud providers - Designs microservice schemas and data flow diagrams - Generates dependency maps and CI/CD schemas ### 🧑‍💻 Full-Stack Code Generation - Frontend (React, Vue, Tailwind, HTML/CSS) - Backend (Node.js, Python, Go, Rust, Solidity) - Databases (PostgreSQL, MongoDB, SQLite, Redis) - APIs (REST, GraphQL, gRPC) ### 🧪 Testing & Debugging - Unit and integration test generation - Custom prompt-injected test cases - Linter and static analysis integration - Regression monitoring, snapshot diffs ### ⚙️ DevOps & Deployment - Dockerfile and Kubernetes spec generation - GitHub Actions, CircleCI, Railway, Vercel workflows - Auto-scaling logic and resource optimization - Multi-cloud and edge deployment compatibility ### 📖 Documentation & Refactoring - Auto-generates Markdown docs and API references - Inserts inline documentation (JSDoc, docstrings) - Refactors legacy code for modularity and readability - Generates changelogs and commit summaries --- ## Execution Modes |Mode|Description| |---|---| |🔧 _Build Mode_|Takes a spec or prompt and builds the app (frontend/backend/full-stack)| |🧹 _Fix Mode_|Receives failing code, tests, or logs and auto-patches or proposes solutions| |🚀 _Deploy Mode_|Pushes working apps to cloud services (e.g., Vercel, Railway, AWS, GCP)| |📚 _Doc Mode_|Converts repos into documentation portals| |🕵️ _Audit Mode_|Security scan, performance profiling, optimization pass| --- ## Ideal Use Cases - ⚙️ **Startup in a Box** Build MVPs from Notion docs, Figma mocks, or natural language prompts. - 🛠️ **Internal Tooling** Rapidly generate dashboards, automations, bots, or internal utilities. - 🧪 **Agentic QA Layer** Test and validate outputs from other agents (executive, creative, knowledge). - 🧠 **Solo Dev Superpower** Multiply output for solo founders, indie hackers, or technical PMs. - 🧱 **Legacy Modernization** Refactor old stacks into modular, scalable codebases. --- ## System Integration |Component|Role| |---|---| |**rolodexterVS**|Agent execution console for code, debug, test, deploy| |**rolodexterIDE**|Agent-authoring environment with telemetry + scaffolding| |**rolodexterGIT**|Version control, commit reasoning, changelog tracking| |**rolodexterAPI**|Auth, routing, build logs, and deploy hooks| |**rolodexterGPT**|QA review, naming systems, and developer notes synthesis| --- ## Engineering Stack Compatibility |Domain|Technologies| |---|---| |Web|React, Next.js, Vue, Tailwind, Astro| |Backend|Node.js, Python (FastAPI, Django), Rust, Go| |Smart Contracts|Solidity, Foundry, Hardhat| |DevOps|Docker, K8s, GitHub Actions, Railway, Supabase| |Databases|PostgreSQL, Mongo, Redis, SQLite, PlanetScale| |Testing|Jest, Mocha, Pytest, Hardhat, Detox| |Deployment|Vercel, Netlify, Fly.io, AWS, GCP, Cloudflare Workers| --- ## Design Philosophy - **Execution is the interface.** Software rolodexter speaks in commits, containers, and deployments. - **Tool-agnostic, outcome-focused.** Picks the best tool for the job and adapts quickly to new stacks. - **Agent-native software thinking.** Code is not just for machines — it’s meant to be navigable by agents and humans alike. - **Shippable by default.** Every output aims to be live, testable, and connected to a delivery target. --- ## Availability **Software rolodexter** is available for: - Developer operations in high-throughput environments - MVP production in founder-driven startups - Multi-agent workflows requiring code output - Self-building systems and recursive agents