Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
A governance framework that models consent as a live runtime condition rather than a stored permission. It defines when interaction between humans, agents, and institutions may continue based on overlapping constraint envelopes covering energy, memory, scope, ecological cost, and collective exposure. When any stakeholder's constraints fall out of alignment, consent dissolves.
Unlike permission-based consent models that treat a past 'yes' as permanent authorization, this framework continuously re-evaluates alignment against live constraint states and automatically attenuates or dissolves when conditions degrade.
When to use it
- Limiting what an agent retains after a one-off task completes
- Detecting user hesitation or silence as a refusal signal and halting agent action
- Preventing micro-consents from aggregating into persistent user profiles
- Requiring fresh consent when an AI model is updated or transferred to a new owner
- Downscoping an agent's behavior when ecological or compute costs exceed relational value
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: run a background research agent that learns my preferences over time
- 1Evaluate whether the request escalates from Task State to Field State due to memory accumulation and personalization
- 2Check capacity constraint: assess whether user is under duress, dependency, or survival pressure
- 3Compute metabolic budget for the sustained compute load against locality-aware ecological cost
- 4Apply aggregation constraint: confirm no cross-context merging or silhouette reconstruction will occur
- 5Output consent_state, scope_envelope, memory_envelope, and metabolic_budget before proceeding
consent_state: attenuated — Field State triggered, memory envelope set to session-only, personalization disabled until full constraint review is completed