A Claude Code Skill for Idea Validation

Your idea has
predators.

Describe your product. 68 AI agents attack it from every angle — market risk, competitive threats, regulatory traps, unit economics. What survives is worth building. What doesn't gets three pivots.

Two commands. No dependencies. No config. Installs in 30 seconds.

GitHub stars License No dependencies
YOUR
IDEA MealBot
PAUSED
0 specimens catalogued
0 chapters written
0 attack categories
0 dependencies required

How It Works

~ three phases of natural selection ~
01

Drop

Your raw idea enters the ecosystem. We sharpen it — ICP, geography, core problem statement, hypothesis.

02

Battle

Animals attack in elimination rounds. Weakest argument dies each round. Survivors absorb the kill insight.

03

Emerge

Battle-hardened idea with unfair advantage statement. Or the idea dies — and three pivots are generated.

Three Outcomes

Every battle ends one of three ways.

Survived

Core thesis held

The idea withstood every attack. Patches were cosmetic. Structural advantage confirmed. Build it.

Restructured

Premise died, tech preserved

The original pitch was killed but core technology survived. New ICP, new moat, new angle emerged.

Killed

Pivot engine activates

No structural advantage found. Three alternative pivots generated and mini-battled for survivability.

Real Battles

Every battle below is real. No parameter tuning between tests.

Viral Content Repurposer
Restructured SUBAGENT
🐦‍⬛ Crow — apex predator
“Creators don’t need speed — they need structural intelligence.”
Original Idea
SaaS tool ($29/mo) for solo creators that detects viral content and auto-repurposes across platforms within minutes. Instagram Reel blows up → YouTube Short, TikTok, Twitter thread, and LinkedIn post are live before the wave peaks.
Ecosystem
🪱 Tapeworm · 🦈 Shark · 🐢 Tortoise · 🦟 Mosquito · 🐦‍⬛ Crow
Key Rounds
R1: Crow destroyed the “catch the wave” premise — platform algorithms are independent, cross-platform speed provides zero advantage. R2: Creator-in-the-loop killed by notification response latency and temporal fingerprinting. R3: Voice personalization exposed as “compounding rigidity” since creators pivot every 6–12 months. R4: Structural optimization creates convergence — platforms detect and suppress templated uniformity at scale.
Evolved Idea
Cross-creator structural intelligence platform. Creator uploads content via share sheet (no API dependency). Tool restructures for each platform using emerging format patterns detected across an anonymized creator cohort — not what worked last month, but what’s working right now.
Unfair Advantage
Anonymized cross-creator performance network detects emerging format patterns 48 hours before public analytics — a signal that compounds with every creator added and cannot be replicated without rebuilding the entire cohort.
AI Meeting Notetaker
Survived FALLBACK
🐦‍⬛ Crow — apex predator
“The meeting bot was a feature, not a product. Survival came from owning the cross-tool org graph.”
Original Idea
AI notetaker ($19/mo per team) for SMEs. Joins meetings via calendar integration, extracts action items, and auto-assigns them in Slack with deadlines. Target: ops managers at growing startups who lose action items because nobody takes notes.
Ecosystem
🪱 Tapeworm · 🐘 Elephant · 🐢 Tortoise · 🦟 Mosquito · 🐦‍⬛ Crow
Key Rounds
R1: Elephant proved Microsoft Copilot and Google Gemini do this for free — bundled with existing licenses. R2: Zoom AI Companion collapses the repositioned niche. R3: Crow reframed the entire product — meetings are the wrong input; Slack threads, email, and Loom comments contain more untracked action items. R4: Neither Crow nor Mosquito delivered a lethal blow; the idea absorbed each reframe and expanded its surface.
Evolved Idea
Omni-channel action capture and routing platform. Extracts action items from meetings, Slack, email, and async video. Pre-seeds organizational graph from Slack workspace data. Routes each item to the right person in the right tool with deadline enforcement and follow-up nudges.
Unfair Advantage
Cross-tool organizational graph — mapping who owns what across Slack, email, meetings, and project tools — compounds with every interaction processed and cannot be replicated by any single-platform player because it requires visibility across ecosystem boundaries they don’t control.
MealBot Office Coordinator
Killed FALLBACK
🐦‍⬛ Crow — apex predator
“70–80% of target offices use contracted caterers, not Swiggy. MealBot automates a workflow that barely exists.”
Original Idea
AI meal coordinator for Indian SMEs (Rs.750–1500/mo). Handles bulk lunch orders from Swiggy/Zomato, tracks employee preferences and dietary restrictions, manages budgets, and auto-orders daily for offices of 10–100 employees in Bangalore, Hyderabad, and Pune.
Ecosystem
🪱 Tapeworm · 🐂 Ox · 🦝 Raccoon · 🐊 Crocodile · 🐦‍⬛ Crow
Key Rounds
R1: Crow delivered a structurally fatal blow — the ICP does not order from Swiggy daily, they use contracted caterers. R2: Narrowed ICP to small startups, but Crow proved growing startups churn into the caterer model within 12–18 months. R3: Caterer management pivot hit a wall — the relationship is optimized for informality, and digitization adds friction. R4: Multi-vendor orchestration already exists — it’s called Sodexo.
Evolved Idea
B2B caterer procurement intelligence for facility managers at 500+ employee companies. Caterer performance benchmarking, contract renewal intelligence, and vendor discovery — a procurement tool, not a meal coordinator.
Unfair Advantage
Cross-company caterer benchmarking network effect: every new client’s caterer ratings, pricing, and hygiene scores improve the dataset, giving facility managers negotiation leverage no single-company tool can replicate.
Email Audio Briefing
Killed FALLBACK
🐦‍⬛ Crow — apex predator
“You started with a megaphone and ended with a magnifying glass, and Google already makes both.”
Original Idea
Browser extension + mobile app ($9/mo) that summarizes unread emails every morning using AI and delivers a 60-second audio briefing. Target: VP/Director level managers at 200–2000 employee companies in US/UK who receive 80–150 emails daily.
Ecosystem
🦂 Scorpion · 🦁 Lion · 🦎 Chameleon · 🦊 Fox · 🐦‍⬛ Crow
Key Rounds
R1: Lion proved Google Gemini and Microsoft Copilot already ship email summaries for free — IT will block a third-party extension before approving it. R2: Cross-platform unification amplified the symptom instead of curing it — the briefing became a firehose narration. R3: Decision-extraction assumed the wrong bottleneck; VPs know which emails need decisions, the problem is making them. R4: Four pivots and zero remaining differentiation — each retreat arrived at features incumbents already shipped.
Evolved Idea
Email thread intelligence API/SDK sold B2B as embedded middleware to existing email clients, CRMs, and project management tools. Not a consumer product — a middleware engine other products embed at $0.002 per email processed.
Unfair Advantage
None survives the battle intact. The original advantage (audio format) died in Round 1. Cross-platform unification died in Round 2. Decision-extraction died in Round 3. This idea has no unfair advantage — proceed only with a technical co-founder who has shipped enterprise email infrastructure before.

The Specimens

~ each animal represents a distinct, documented threat vector ~

Field Reports

~ real battles, real ideas, documented outcomes ~

Live Observation

~ pick an idea, watch the predators attack ~
Observation in Progress DEMO
Select a specimen above to begin observation...
Pre-scripted excerpts from documented field reports. Install the guide for the real thing.

Chapters

~ five skills, one install, complete lifecycle ~
I
food-chain-ideation
Product Idea Battles

35 animals attack your product idea in elimination rounds. Blind scoring. Audience simulation. Pivot engine when killed.

Trigger: food chain my idea: ...
II
food-chain-code
Architecture Battles

21 code animals stress-test your architecture. Scale, security, maintenance, and team-dynamics attacks.

Trigger: food chain code: ...
III
food-chain-pitch
Pitch Battles

12 pitch animals attack your fundraise narrative. Unit economics, due diligence, market sizing, founder-market fit. Investor simulation panel.

Trigger: food chain pitch: ...
IV
apex-to-action
Battle → 90-Day Plan

Converts battle output into a validated 90-day execution plan. Validation gates. Three phases. Kill list.

Trigger: apex to action: [paste log]
V
food-chain-monitor
Delta Re-Testing

After pivots or market changes, tests only new assumptions. 3 animals, 2 rounds max. Under 5 minutes.

Trigger: food chain monitor: [changes]

Begin your field work.

zero dependencies, works everywhere
1. Add from marketplace
claude plugin marketplace add CodedRichy/food-chain-ideation
2. Install the skill
claude plugin install food-chain-ideation@food-chain-ideation
Works with Claude Code CLI and Claude.ai
Clone into your project
git clone https://github.com/CodedRichy/food-chain-ideation.git
Copy skills into .cursor/skills/
cp -r food-chain-ideation/skills .cursor/skills
Skills load automatically from .cursor/skills/ in your project root.
Clone into your project
git clone https://github.com/CodedRichy/food-chain-ideation.git
Copy skills into .windsurf/skills/
cp -r food-chain-ideation/skills .windsurf/skills
Skills are markdown-based — no build step, no dependencies.
Clone into your project
git clone https://github.com/CodedRichy/food-chain-ideation.git
Copy skills into .github/skills/
cp -r food-chain-ideation/skills .github/skills
Reference skill files in your Copilot instructions or .github/copilot-instructions.md.
Clone the repo
git clone https://github.com/CodedRichy/food-chain-ideation.git
Copy to your editor's skill directory
cp -r food-chain-ideation/skills <your-editor>/skills
Pure markdown — no config, no build step, no dependencies.
Common paths: .claude/skills/ · .cursor/skills/ · .windsurf/skills/ · .github/skills/
On Windows, use copy or xcopy instead of cp -r.
Coming soon
In the field
food-chain-hire

Hiring decision battles

In the field
food-chain-price

Pricing model attacks