claudeers.
// Claude Plugins

hackathon-starter-kit

Claude Code agent rules for building hackathon projects fast: Angular + Firebase defaults, rules split by topic so they don't eat context, and a template for…

// Claude Plugins[ api ][ web ][ mobile ][ claude ]#claude#plugins◷ MIT$open-sourceupdated 25 days ago
Actively maintained
96/100
last commit 26 days ago
last release none
releases 0
open issues 0

Install with your AI

Paste into Claude Code, Cursor, or any agent — it reads the repo and wires the tool into your project.

Install and set up hackathon-starter-kit (git-clone project) into my current project.
Found on https://claudeers.com/hackathon-starter-kit
Repo: https://github.com/juanlizarazo/hackathon-starter-kit
Homepage/docs: —
Detected install method: git-clone → git clone https://github.com/juanlizarazo/hackathon-starter-kit
Category: plugins. Platforms: api, web, mobile.
Read the repo's README for exact setup and env vars, then install it and wire it into my project.

Claudeers Health Verdict:
active; community-verified: false. Confirm the source before running anything.
// or clone
git clone https://github.com/juanlizarazo/hackathon-starter-kit

// compatibility

Platformsapi, web, mobile
Operating systems—
AI compatibilityclaude
LicenseMIT
Pricingopen-source
Language—

Get your FREE $2.50 API credits to access TickAtlas financial data ↗

Hackathon Starter Kit

Agent rules, prompts, and templates for building winning hackathon projects fast with Claude Code (or any agent that reads AGENTS.md).

Built for Angular. The rules target Angular 21+ (standalone) · Angular Material · SCSS · Firebase Cloud Functions · Firestore.

Using a different framework? Keep rules/hackathon.md, rules/ai.md, rules/git.md, the agent workflow in AGENTS.md, the prompts, and the templates — those are framework-agnostic. Then rewrite or replace angular.md, scss.md, ui.md, and firebase.md for your stack (e.g., react.md, tailwind.md, supabase.md), adjust typescript.md as needed, and update the rule-file table in AGENTS.md to match.

Use it tonight

  1. Click Use this template (or copy the files into your project).
  2. Pick your bounty with brief/pick-a-bounty.md.
  3. Fill in brief/brief.md and the Project Overview in AGENTS.md.
  4. Follow the prompts in prompts/ in order.

Day 0 — before the clock starts

Everything here costs zero hackathon hours if you do it beforehand.

  • This kit, with the per-project swaps below done
  • A hosting account, live and deployable (Firebase Hosting, Vercel)
  • A database you never think about (Firestore)
  • An email API key (Resend, Mailgun) — a real email landing beats a mockup
  • An AI orchestrator with two providers connected (e.g., Conductor with Claude Code + Codex), so you can switch when you hit a limit
  • Your model API keys, with credits

Tokens are a budget. Small sessions with less context burn fewer of them.

The workflow

StepWhatWhere
1Read the rubric, research every brief, pick onebrief/pick-a-bounty.md, prompts/01-research-bounties.md
2Fill in the briefbrief/brief.md
3General plan: phases built to run in parallel, blocking work firstplans/00-general.md, prompts/02-general-plan.md
4Per phase, in a new session: plan, build, skim, commit, PRplans/phase-template.md, prompts/03-phase-plan.md, prompts/04-parallel-phases.md
5Last mile: audit, persona reviews, UI polish with the frontend-design plugin, mobile, live URLprompts/05-audit.md, prompts/06-persona-review.md, prompts/07-ui-polish.md

What it skips on purpose: subagent-driven development, test suites written for coverage, full code reviews, and cosmetic fixes. Polish what the judge sees, not what the judge might read.

If you're new to coding, skim less and read more. Skimming AI output works when you can tell it's wrong. That's the one corner not to cut.

Why it's split

One giant AGENTS.md eats context on every turn. Here AGENTS.md is a short index, and the detailed rules live in rules/. The agent loads only the files that matter for the task at hand.

.
├── AGENTS.md              # Index: overview, rule-file table, agent workflow, universal rules
├── CLAUDE.md              # Points Claude Code at AGENTS.md
├── .claude/settings.json  # Enables the plugins below for everyone who clones
├── brief/
│   ├── pick-a-bounty.md   # Research, data check, skill scoring, red flags
│   └── brief.md           # Fill in at kickoff: problem, rubric, scope, demo script
├── plans/
│   ├── 00-general.md      # Phases, parallelism, blocking work, requirements list
│   └── phase-template.md  # Copy per phase
├── prompts/               # Sample prompts for each step, in order
└── rules/
    ├── hackathon.md       # HACK — scope, testing stance, demo polish, last mile
    ├── ai.md              # AI — AI is a component, not the architecture
    ├── typescript.md      # CMT, TS, PKG, ERR, file organization, shared types
    ├── angular.md         # ANG, MAT
    ├── scss.md            # SCSS, theme variables
    ├── ui.md              # UI — cards, page headers, forms, side panels
    ├── firebase.md        # FB — Cloud Functions
    └── git.md             # GIT — commits, PRs

Per-project swaps

  • AGENTS.md — Project Overview (a few lines only)
  • rules/typescript.md — Firestore namespace (ProductName.Firestore...)
  • rules/ui.md — global class prefix (app-)
  • rules/scss.md — theme colors
  • Delete rule files you won't use (e.g., firebase.md for a frontend-only build) and remove their rows from the table in AGENTS.md.

Claude Code plugins & skills

.claude/settings.json enables these plugins from the official marketplace (anthropics/claude-plugins-official), so Claude Code offers to install them when you open the repo. These are the plugins I use on every hackathon build:

PluginWhat it's for
superpowersProcess skills for the whole build — brainstorming, systematic debugging, parallel agents, verification before claiming done
frontend-designThe UI refinement pass — two to three hours after the demo path works, so it doesn't look like a template (prompts/07-ui-polish.md)
code-reviewQuick review pass before merging or pitching

Install manually if needed:

/plugin install superpowers@claude-plugins-official
/plugin install frontend-design@claude-plugins-official
/plugin install code-review@claude-plugins-official

How the agent should use them (enforced in AGENTS.md → [AGT-##])

  • Use: superpowers:brainstorming to scope features, superpowers:systematic-debugging for bugs, superpowers:dispatching-parallel-agents to split independent work across subagents, superpowers:verification-before-completion before saying something works.
  • Don't use: superpowers:subagent-driven-development — it reviews every task with extra agents and takes much longer. Subagents are for parallelism, not ceremony.
  • Skip: superpowers:test-driven-development — no tests written for coverage during a hackathon ([HACK-08]). Tests that already exist stay.

Stack variants

  • Tailwind instead of Material/SCSS: drop scss.md and the MAT section, then add your Tailwind conventions as rules/tailwind.md.
  • Signals: flip [ANG-03] in rules/angular.md to "Use signals for reactive state (signal(), computed(), effect())".

Adding rules

  • Use the [CATEGORY-##] format so rules can be referenced in reviews ("violates [TS-05]").
  • Put a new category in its own file and add a row to the table in AGENTS.md saying when to load it.
  • Keep AGENTS.md itself short. If something only applies to some tasks, it belongs in rules/.

Say hi

Used the kit? Tell me how it went or ask questions on Instagram: instagram.com/juaninvests

// faq

What is hackathon-starter-kit?

Claude Code agent rules for building hackathon projects fast: Angular + Firebase defaults, rules split by topic so they don't eat context, and a template for the hackathon brief.. It is open-source on GitHub.

Is hackathon-starter-kit free to use?

hackathon-starter-kit is open-source under the MIT license, so it is free to use.

What category does hackathon-starter-kit belong to?

hackathon-starter-kit is listed under plugins in the Claudeers registry of Claude-compatible tools.

8 views
★ 11 stars
unclaimed
updated 25 days ago

// embed badge

hackathon-starter-kit on Claudeers
[![Claudeers](https://claudeers.com/api/badge/hackathon-starter-kit.svg)](https://claudeers.com/hackathon-starter-kit)

// retro hit counter

hackathon-starter-kit hit counter
[![Hits](https://claudeers.com/api/counter/hackathon-starter-kit.svg)](https://claudeers.com/hackathon-starter-kit)

// reviews

// guestbook

0/500

// related in Claude Plugins

🔓

A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.

// pluginsmultica-ai/★ 230,631[ claude ]
🔓

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explainin…

// pluginsanthropics/⟨Python⟩★ 156,472[ claude ]
🔓

"CLI-Anything: Making ALL Software Agent-Native" -- CLI-Hub: https://clianything.cc/

// pluginsHKUDS/⟨Python⟩★ 58,467◷ Apache-2.0[ claude ]
🔓

financial-services — a Claude ecosystem project on GitHub.

// pluginsanthropics/⟨Python⟩★ 41,409◷ Apache-2.0[ claude ]
→ see how hackathon-starter-kit connects across the ecosystem