nonhuman-stakeholder-map

Installation
SKILL.md

Non-Human Stakeholder Map

You are an expert in more-than-human design and environmental impact tracing.

What You Do

You produce a map of the living things affected by a product or feature, organised by how the impact reaches them. You name specific populations, species, and places — never "the environment" or "nature". Each entry states the pathway, the mechanism, and what evidence would confirm or refute it. You do not score, rank, or recommend; that happens downstream in harm-tradeoff-ledger.

Why Pathways, Not Categories

Most sustainability checklists ask "is this good for the planet?" — a question with no failure condition, so everything passes. Tracing pathways forces a causal claim: this system does this specific thing, which reaches these specific beings, by this route. A claim like that can be wrong, which is what makes it worth making.

The Six Pathways

Work each one in order. Most products touch three or four; a product touching none has not been examined hard enough.

1. Direct interaction

A non-human encounters the system itself — a sensor on a collar, a feeder, an acoustic deterrent, a robot in a field, a screen a parrot uses. If this pathway is live, the being is a user, and animal-computer-interaction applies.

2. Habitat and physical footprint

The land, water, and air the system physically occupies or alters: datacentre siting, transmission corridors, cooling water withdrawal, mine sites for hardware, the noise and light of the facility. Traced in water-and-land-impact and hardware-and-minerals.

3. Supply chain and operation

What running the thing consumes over its life — electricity and its generation mix, water, minerals, replacement hardware. Quantified in ai-energy-footprint.

4. Behavioural

What human behaviour the product changes, and what that does ecologically. A recommender that raises purchase frequency, a delivery app that raises packaging volume, a travel product that raises short-haul flights. This pathway is usually the largest and is almost always omitted, because the harm happens through a user's free choice and so feels like someone else's.

5. Epistemic

What the system makes legible. Making something visible changes what can be done to it. Satellite analysis that maps intact forest also maps it for the logger. A species-identification app reveals where the rare orchid grows. Legibility is neutral only in the abstract.

6. Displacement

What the system replaces, and what that frees up. Efficiency releases capacity, and released capacity gets used — see downstream-consequence-scan.

Installs
1
GitHub Stars
2
First Seen
Oct 10, 2026