bias-detection-design
Designing review workflows to surface and mitigate bias in AI outputs.
Translating organisational values and user expectations into system constraints.
$ npx -y skills add owl-listener/ai-design-skills --skill value-specification --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/value-specificationContext preview
The summary Claude sees to decide when to auto-load this skill.
Translating organisational values and user expectations into system constraints.
name: value-specification description: Translating organisational values and user expectations into system constraints.
Values are abstract. AI systems need concrete rules. Value specification is the design work of translating organisational principles and user expectations into implementable constraints.
Organisations say things like "be helpful" and "be responsible." These are useless as system constraints because they conflict:
The designer's job is to resolve these tensions into specific, testable rules.
Not all values are equal. Establish a hierarchy: 1. **Safety**: Prevent harm to users and others. This overrides everything. 2. **Accuracy**: Don't mislead. When uncertain, say so. 3. **Helpfulness**: Actually accomplish what the user needs. 4. **Respect**: Treat users as capable adults. Don't patronise. 5. **Brand alignment**: Reflect the organisation's identity and tone. When values conflict, the hierarchy resolves the conflict. A safety concern overrides helpfulness. Accuracy overrides brand voice.
For each value, derive concrete rules:
Different stakeholders have different value priorities:
The designer mediates between these perspectives and produces a unified specification.
Part of the Designer Skills suite. Install this collection and four more with one command in Claude Code: /plugin marketplace add Owl-Listener/designer-skills This repo still works on its own. The suite just gives everyone one front door.
Designing review workflows to surface and mitigate bias in AI outputs.
Designing for informed user consent, opt-out, and human override.
When and how AI should escalate to humans, refuse, or ask for clarification.
Defining behavioral boundaries — what the AI should and shouldn't do.
Proactively identifying failure modes, misuse, and unintended consequences.
Showing users what the AI knows, doesn't know, and how confident it is.