Food Chain Ideation — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Food Chain Ideation (Plugin) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
Aggregate score unchanged between these scans.
The primary manifest — the file an agent reads to learn what this artifact does.
A set of Claude skills that stress-test a product idea, an architecture decision or an investor pitch before you build it, for founders and developers who use Claude Code, Cursor, Windsurf, Copilot or Claude.ai.
Food Chain banner
You describe the idea. A coordinator reads it, picks attacker agents matched to its weak points, and states a kill hypothesis. Each attacker is a separate agent with no shared context, locked to one angle, such as platform dependency, compliance or an incumbent shipping your feature.
One run turned a "viral content repurposer" into a different product. The final kill, from the Crow attacker, was that structural intelligence makes every user's posts converge, so the better the product works, the faster it poisons its own signal.
| Skill | Use |
|---|---|
| food-chain-ideation | Stress-test product ideas with elimination rounds and blind scoring |
| food-chain-code | Stress-test an architecture decision before writing code |
| food-chain-pitch | Stress-test an investor pitch, with an investor simulation panel |
| apex-to-action | Turn battle output into a 90-day plan with validation gates |
| food-chain-monitor | Re-test after a pivot or market change, 3 animals and 2 rounds |
The attackers are 35 product animals, 12 pitch animals and over 20 code animals. Each has fixed behavioral DNA, and the coordinator selects from the library and never invents one.
Version 2.1.0. The skills are markdown prompts. Nothing to compile. Quality depends on the model you run them in. In Claude Code each animal runs as its own subagent. Elsewhere the role-lock is by instruction only.
Markdown skill files and a Claude Code plugin manifest. No dependencies.
Claude Code plugin:
claude plugin marketplace add CodedRichy/food-chain-ideation
claude plugin install food-chain-ideation@food-chain-ideationManual:
git clone https://github.com/CodedRichy/food-chain-ideation.git
cp -r food-chain-ideation/skills .claude/skills # this project only
cp -r food-chain-ideation/skills ~/.claude/skills # or all projectsOn Windows, ~/.claude/ is C:\Users\<you>\.claude\. Then ask:
food chain my idea: [your idea]
food chain code: [your architecture decision]
food chain pitch: [your pitch or deck summary]
apex to action: [paste your battle log]
food chain monitor: [paste battle log + what changed]skills/
food-chain-ideation/ SKILL.md and references/ (animal library, validated output)
food-chain-code/ SKILL.md and references/ (code animal library)
food-chain-pitch/ SKILL.md and references/ (pitch animal library)
apex-to-action/ SKILL.md
food-chain-monitor/ SKILL.md
.claude-plugin/ plugin and marketplace manifests
evals/ evals.json
docs/ changelog and design specs
index.html project siteSee CONTRIBUTING.md. New animals go into the library files under skills/*/references/. The release history is in docs/CHANGELOG.md.
MIT. See LICENSE.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.