跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
其他技能

think-ethical-matrix

Produces an ethical matrix, a grid that cross-references affected parties (rows, deliberately including voiceless ones such as future generations and the environment) against prima facie principles (columns of wellbeing, autonomy, fairness), specifying the option's impact in every cell and reading the result as a trade-off pattern rather than a score. Use when a concrete proposal affects multiple parties with a genuine moral trade-off among them, some affected parties have no voice, and an ethics debate keeps sliding between groups and principles. Not a decision calculator and not an option-versus-criteria scorer; the matrix maps the terrain and leaves the weighing to deliberation.

product-on-purpose · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add product-on-purpose/thinking-framework-skills --skill think-ethical-matrix

技能说明


name: think-ethical-matrix description: Produces an ethical matrix, a grid that cross-references affected parties (rows, deliberately including voiceless ones such as future generations and the environment) against prima facie principles (columns of wellbeing, autonomy, fairness), specifying the option's impact in every cell and reading the result as a trade-off pattern rather than a score. Use when a concrete proposal affects multiple parties with a genuine moral trade-off among them, some affected parties have no voice, and an ethics debate keeps sliding between groups and principles. Not a decision calculator and not an option-versus-criteria scorer; the matrix maps the terrain and leaves the weighing to deliberation. license: Apache-2.0 metadata: id: thinking-framework-skills.ethical-matrix family: ethics-values-deliberation evidence-tier: "P" version: 0.1.0 standard: "0.8"

Ethical Matrix

An "is this ethical?" debate over a concrete proposal tends to slide: it starts on one affected group, drifts to a principle, jumps to a different group, and never holds both axes at once - so the trade-offs that actually matter (a benefit to one party paid for by a burden on another, on a different principle) stay invisible. The ethical matrix refuses that slide. It cross-references the affected parties (rows) against a small fixed set of prima facie ethical principles (columns), and forces a concrete impact specification in every party-by-principle cell. The rows deliberately include parties who cannot speak for themselves - animals, ecosystems, future generations, absent groups - because they get a row whether or not anyone in the room represents them. The durable move is principled cross-referencing: making the trade-offs a single-axis analysis hides become visible and individually contestable.

Two design choices are load-bearing and frequently misunderstood. First, the matrix maps the moral terrain; it does not weigh it. The grid is not an algorithm and emits no verdict - the weighing happens in deliberation over the filled grid. Second, the cells are impact specifications ("how does this affect the wellbeing of small producers"), not perspective voicings ("what would small producers say"). That second point is what separates it from stakeholder perspective-taking. The output is an ethical matrix: a party-by-principle impact grid with each cell tagged factual or contested, a trade-off pattern read-out, and an explicit no-verdict footer.

When to Use

  • A concrete proposal, technology, policy, or feature has multiple affected parties and a genuine moral trade-off among them: who gains, who pays, and on what dimension.
  • Some affected parties have no voice - non-human, future, or absent groups that need representing even though nobody in the room speaks for them.
  • An ethics debate keeps sliding between groups and principles without anyone noticing the slide, and disagreement needs to attach to specific cells rather than to vague unease.
  • A group needs a shared, inspectable structure for a values trade-off, so a position can be defended cell by cell.

When NOT to Use

  • Do not present it as a decision calculator. The matrix cannot weigh the problems it uncovers (Schroeder and Palmer, 2003: helpful for fact-finding, "much less helpful" for weighing). A filled grid presented as an answer is the method's signature abuse. This is the central wall.
  • Do not use it to score options against the decider's own criteria. That is think-decision-option-review (options by weighted criteria, aggregated to a recommendation). The ethical matrix maps one proposal's impacts on affected parties against impartial principles and refuses to aggregate.
  • Do not use it when who counts as affected is the live dispute. The matrix takes its row roster as given. If frame membership is contested, audit the boundary first with think-boundary-critique, then populate the rows.
  • Do not impose the principle columns unreflectively. The fixed principlist columns can crowd out the values participants actually hold (Cotton, 2009). The columns may be adapted, but adaptation must be deliberate and stated, not silent.
  • Do not run it as completeness theater. A fully populated grid looks rigorous while the substantive judgments hide inside cell wording. The cells are claims to challenge, not boxes to tick.

Instructions

When asked to assess the ethics of a concrete proposal across affected parties, follow these steps:

  1. State the option in one line. Name the single concrete proposal, technology, policy, or feature under analysis. The matrix evaluates one option's impacts; it is not a comparison of options (that is think-decision-option-review).
  2. Roster the affected parties (rows). List the groups the option affects. Deliberately include parties who cannot speak for themselves - animals, ecosystems, future generations, absent or unrepresented groups. Each gets a row whether or not anyone present represents them. If who counts as affected is itself the contested question, stop and run think-boundary-critique first.
  3. State the principle columns. Use the default prima facie set - wellbeing (beneficence and non-maleficence together), autonomy, and fairness (justice). If the situation needs an adapted set (for example a future-generations or solidarity column for an algorithmic-audit context), state the adaptation and justify it explicitly. Never swap the columns silently. When the parties' moral concerns differ in kind, not just in degree, an optional moral-foundations column set can replace or supplement the default three. MFQ-2 names six foundations: care, equality, proportionality, loyalty, authority, and purity (Atari et al., 2023). Use it only to name each party's concern, not to score it, and state the swap as the adaptation this
阅读完整来源 ↗