Cybernetics

Bio-Cybernetics Review

Judge whether a thing will stay alive. Living systems that survive over time share a small set of design principles; Frederic Vester distilled them into 8 bio-cybernetic rules. This skill treats the subject — an idea, product, business model, org, policy, or architecture — as a living system embedded in a larger one, maps its feedback loops, and scores it against the 8 rules. A subject that honors them tends to be resilient and self-sustaining; one that breaks them tends to grow brittle, waste energy, or crash when a higher-order loop intervenes.

> The 8 rules below are Frederic Vester's biocybernetic rules (from Die > Kunst, vernetzt zu denken / "The Art of Interconnected Thinking"). This skill > turns them into a review and improvement diagnostic.

What this skill produces

Given a subject, output a structured review: the subject as a system → its feedback map → a per-rule scorecard → the highest-leverage fixes → an overall viability verdict → signals to watch. Fill the [output template](#output-template).

For deep rule cards with Vester's original examples, violation signals, and a worked example, read references/biocybernetics-field-guide.md when a call is non-obvious or you want the source illustrations.

How cybernetics differs from systems-thinking

They compose well but answer different questions. systems-thinking classifies a problem (clear/complicated/complex/chaotic) and picks the right action protocol. cybernetics evaluates the viability of a system's design against 8 specific bio-inspired rules and proposes redesigns. Reach for this skill when the question is "will this survive and stay healthy over time, and how do I make it more so?"

Step 1 — Frame the subject as a living system

A bio-cybernetic review only works once the subject is stated as a system. In one line each:

• Subject — what it is.
• Function — the need it actually serves (not the product/form it currently takes — this matters for Rule 3).
• Enclosing system — the higher-order system it lives inside (its market, ecosystem, body politic, environment). Rule 1 turns on this: when a subsystem's own brakes fail, a higher regulating circuit intervenes — often by killing it.

If you can't name the function and the enclosing system, get those before scoring.

Step 2 — Map the feedback loops

Feedback is the spine of cybernetics (Rules 1 and 8 are explicitly about it). Name the loops driving the subject's behavior:

• Reinforcing (positive) loops — self-amplifying: more → more (growth, virality, accumulation, arms races).
• Balancing (negative) loops — self-regulating: deviation triggers correction (saturation, price, fatigue, quotas, predators). These are the governors.
• Dominant loop today — which one actually drives current behavior, and what happens if it keeps dominating.

Step 3 — Score the 8 rules

For each rule give a verdict, the evidence, and the leverage move. Use the scale: ✓ Aligned · ⚠ At risk · ✗ Violated · – N/A. Don't force all eight to apply — mark N/A honestly; the value is in the few that bite.

| # | Rule | What to check | Violation signal | Leverage move | | --- | --------------------------------------------------- | -------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------- | | 1 | Negative feedback dominates positive | Do balancing loops cap the reinforcing ones? Is the regulator part of the system? | Unbounded growth; "more is always better"; no circuit breaker; success accelerates with no brake. | Build in self-limiting loops (quotas, price signals, saturation, rest cycles) before a higher loop crashes it. | | 2 | Vitality independent of quantitative growth | Does survival depend on perpetual growth? Has added size/connectivity started to reduce stability? | Growth is the goal; metamorphosis avoided; ever-more connections breeding fragility or chaos. | Switch from growing to developing — metamorphose; find the optimal, not maximal, size and connectivity. | | 3 | Function-oriented, not product-oriented | Is identity bound to a specific product/form, or to the function it serves? | "We are our product"; can't imagine surviving the product's obsolescence. | Reframe around the job-to-be-done; dare to question the product itself (the VW Beetle won on global service). | | 4 | Use existing forces (jiu-jitsu, not boxing) | Is the system spending energy fighting forces it could ride? | Brute-force control against gradients (air-conditioning the heat you generated); fighting momentum head-on. | Redirect existing flows — sun/wind/waste-heat, market momentum, people's existing motivation — for low steering energy. | | 5 | Multiple use of products, functions, structures | Do elements each serve one purpose where they could serve several? | Single-purpose components; siloed structures; "one tool, one job" everywhere. | Combine functions (combined heat-and-power, dual-use parts, one team many roles) — two birds, one stone. | | 6 | Recycling — use waste and byproducts | Are waste/byproduct streams discarded instead of fed back? | Linear take-make-dispose; outputs leave as loss; "waste" treated as inevitable. | Close loops — every output is some process's input. Nature has no waste; every product has its enzyme. | | 7 | Symbiosis — mutual use of difference | Could complementary partners couple and exchange to save energy/resources? Is it needlessly autarkic? | Doing everything alone; ignoring partners whose differences complement; centralizing where exchange would save. | Form symbioses — couple with the different (the mitochondrion's bargain); prefer many small local exchanges. | | 8 | Biological design via feedback planning | Is the design compatible with humans and nature? Does it connect things without structure, propagating faults? | Connect-everything-to-everything with no insulation; designs hostile to people/environment; plan-then-ship with no feedback. | Design iteratively with feedback; add structure/insulation so interference doesn't spread; keep it human- and nature-compatible. |

Step 4 — Find the highest-leverage fixes

Cybernetics is about leverage: where does a small change yield a large, system-wide improvement? Of the rules scored ⚠ or ✗, pick the 2-3 that, if fixed, most improve viability — usually the ones touching the dominant loop from Step 2. Rank them; don't dilute the report with all eight.

Step 5 — Render the viability verdict

Synthesize one judgment of whether the subject will survive and stay healthy, and name the dominant failure path if nothing changes. Useful labels:

• Self-sustaining — balancing loops govern; honors most rules.
• Brittle growth — alive only while it grows; metamorphosis overdue (Rule 2).
• Runaway risk — a reinforcing loop with no governor; a higher loop will eventually correct it hard (Rule 1).
• Energy-wasteful — fights forces it could ride; leaks waste it could recycle (Rules 4, 6).
• Isolated — autarkic where symbiosis would save it (Rule 7).

Step 6 — Name the signals to watch

For each top risk, give one leading indicator that would confirm or refute the read early — the cheap measurement that reveals whether the missing loop is actually biting (e.g., for Rule 1, track the quantity that's growing unchecked, not the headline that looks healthy).

Output template

## Subject
 — function:  — embedded in: 

## Feedback map
- Reinforcing (positive) loops: ...
- Balancing (negative) loops: ...
- Dominant loop today: 

## Bio-cybernetic scorecard
| # | Rule | Verdict | Evidence | Leverage move |
|---|------|---------|----------|---------------|
| 1 | Negative feedback dominates             | ✓/⚠/✗/– | ... | ... |
| 2 | Vitality independent of growth          | ✓/⚠/✗/– | ... | ... |
| 3 | Function- not product-oriented          | ✓/⚠/✗/– | ... | ... |
| 4 | Use existing forces (jiu-jitsu)         | ✓/⚠/✗/– | ... | ... |
| 5 | Multiple use                            | ✓/⚠/✗/– | ... | ... |
| 6 | Recycling                               | ✓/⚠/✗/– | ... | ... |
| 7 | Symbiosis                               | ✓/⚠/✗/– | ... | ... |
| 8 | Biological / feedback design            | ✓/⚠/✗/– | ... | ... |

## Highest-leverage fixes
1. Rule  — 
2. Rule  — ...

## Viability verdict
 —

## Watch signals
-  → leading indicator: ...

Notes

• Keep the review proportional to the stakes — a small idea gets a few lines and its 2-3 live rules, not the whole template.
• The point is not a high score; it's finding the one or two loops whose redesign changes the trajectory. A subject can violate several rules and still be saved by fixing the dominant one.
• Rules 1 and 2 are about staying viable; 3 keeps you flexible; 4-7 are about efficiency through connection (forces, multi-use, recycling, symbiosis); 8 is about designing with feedback and fit. If short on space, lead with whichever cluster the subject most violates.
• If essential parts are unknown (the function, the enclosing system, the real loops), say what you'd need to observe rather than guessing.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: titusz
• Source: titusz/skills
• License: MIT

Install and usage instructions live in the source repository linked above.