Install
git clone https://github.com/Spirizeon/hackfishInstall command taken from the project's README or registry entry. Always review code before running it.
README excerpt
Read full README on GitHub →Hackfish — Hackathon Simulation System
An AI agent system that runs hackathon simulations to predict winners based on 260+ real hackathon patterns.
Overview
The system simulates a complete hackathon using AI agents (participants, mentors, judges) that interact, make mistakes, learn, and build projects — exactly like real humans. The simulation runs for 48 ticks (one tick = one hour of hackathon time).
Quick Start — Copy-Paste Prompts
Run a full simulation with 800 participants, verbose display, memory storage, and build guide:
run a hackathon simulation skill for [HACKATHON_URL], display, log and store individual agentic memories. total participants 800 teams
Run with build plan generation:
simulate hackathon URL: [URL], 800 participants, display logs, store memories in memories/, generate build guide
Scale to 1000 participants:
predict who will win [HACKATHON_URL] with 1000 participants, store all agent memories, output build plan
Table of Contents
| Section | Description |
|---|---|
| Overview | Project summary, quick start prompts, key features |
| File Structure | Directory layout for simulations |
| 48-Tick Simulation Flow | Phase 1 (Discuss), Phase 2 (Verify), Phase 3 (Develop) |
| Agent Types & Behavior | Participants, Mentors, Judges details |
| Channels Directory | Team chat, mentor help, judge Q&A logs |
| Memory Logging Rules | Individual agent memory files |
| Winning Patterns | 260+ hackathon analysis (sponsor APIs, AI agents, etc.) |
| Scoring Algorithm | Novelty, Feasibility, Impact, Differentiation, Sponsor Bonus |
| How to Run a Simulation | Method 1 (Skill), Method 2 (Python script) |
| Example Simulation Output | Sample results display |
| Technical Details | Agent communication matrix, error handling |
| Files Explained | SKILLS.md, AGENTS.md, README.md |
| Key Principles | Rules: NO CODING until Phase 3, mentor verification, etc. |
| Verbose Display Rules | How agents must display actions during simulation |
Key Features
- 48-tick simulation with realistic agent behavior
- Individual memory logging for every agent
- Dedicated channels/ for team chats, mentor help, and judge Q&A
- 3-phase flow: Discuss → Verify with Mentors → Develop
- Continuous mentor verification between every tick (1-32)
- Mistakes & learning: Juniors struggle, ask for help, seniors guide
File Structure
Similar plugs
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
An AI skill that provides design intelligence for building professional UI/UX across multiple platforms.
Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local…
Why use many token when few token do trick. Viral skill + proxy for coding agents that cuts 65% of tokens by talking like a caveman.
Production-grade engineering skills for AI coding agents.