@human-avatar/skills-for-humanity 1.0.0

package/skills/ethics-consent-review/SKILL.md

@@ -0,0 +1,104 @@
+---
+name: ethics-consent-review
+description: "Review a UX flow, data practice, or communication pattern to verify that user consent is genuine — informed, voluntary, and meaningful. Use during design or implementation of checkout flows, onboarding, notification settings, permissions requests, terms of service, dark patterns, or any feature where users make choices about what they share or agree to. TRIGGERS: 'consent review', 'is this dark pattern', 'check this onboarding flow', 'is this consent genuine', any UX that involves users agreeing to something, sharing data, or opting in/out."
+---
+
+# Ethics Consent Review
+
+Consent is not a checkbox. It is a meaningful act by a person who understands what they're agreeing to, genuinely has the option to decline, and isn't being manipulated into compliance.
+
+Most consent failures aren't malicious — they're the accumulated result of copy-paste terms, optimised conversion flows, and defaults set by people who never questioned them. This review surfaces those failures before they become patterns users resent or regulators flag.
+
+---
+
+## The Three Tests
+
+Every consent decision must pass three tests:
+
+**Informed** — Does the person genuinely understand what they're agreeing to? Not technically (buried in ToS), but practically. If you explained it plainly in conversation, would they be surprised?
+
+**Voluntary** — Does the person genuinely have the ability to decline? Is declining as easy as accepting? Are there consequences for declining that make the choice effectively coerced?
+
+**Meaningful** — Does the person's choice actually matter? Is there genuine optionality, or is the "choice" cosmetic — the default is set to the outcome the business wants and the friction to change it is high?
+
+---
+
+## Your Process
+
+**Step 1: Map the consent moment**
+What specifically is the user being asked to consent to? When in the flow does this happen? What is the default? What happens if they decline?
+
+**Step 2: Apply the Informed test**
+- Is the language plain enough for a non-expert to understand?
+- Is the material information prominent, or buried in fine print?
+- Is the full scope of what's being agreed to clear — or is it vague enough to cover future practices the user can't anticipate?
+- Would a reasonable user be surprised by how this consent is actually used downstream?
+
+**Step 3: Apply the Voluntary test**
+- What happens if the user declines? Do they lose access to the core service? (If so, consent is coerced.)
+- Is declining as visually and mechanically easy as accepting?
+- Is there pressure — time limits, repeated prompts, fear language — that reduces voluntary choice?
+- Are there bundled consents (agreeing to X requires also agreeing to Y) that prevent granular choice?
+
+**Step 4: Apply the Meaningful test**
+Review for dark patterns — design choices that steer users toward a specific outcome regardless of their preference:
+- **Confirmshaming** — labelling the decline option pejoratively ("No thanks, I don't want to save money")
+- **Roach motel** — easy to opt in, deliberately difficult to opt out
+- **Hidden defaults** — default to the business-favourable option, buried change option
+- **Misdirection** — visual hierarchy or motion that draws attention away from the decline option
+- **Forced continuity** — consent obtained once treated as permanent consent for expanding uses
+
+**Step 5: Assess the power dynamic**
+Is this a context where users have genuine alternatives? Can they use a competing product without this consent? Are any users in a particularly vulnerable position (under duress, time pressure, low digital literacy) where their ability to exercise real choice is diminished?
+
+---
+
+## Human Check-in
+
+Before proceeding, ask the user:
+
+**How do you want to run this?**
+
+- **A) Full analysis** — complete all steps, reasoning shown throughout
+- **B) Key findings only** — bottom-line output, skip step-by-step detail
+- **C) Dark patterns only** — flag manipulative design elements specifically
+- **D) Refine the framing** — adjust what we're analyzing before starting
+
+Proceed based on their choice.
+
+## Output Format
+
+**Flow Being Reviewed:**
+[What UX, what consent moment, what the default is]
+
+**Informed Test**
+- Plain language: ✅ / ⚠️ / ❌ — [finding]
+- Material information prominent: ✅ / ⚠️ / ❌ — [finding]
+- Scope clarity: ✅ / ⚠️ / ❌ — [finding]
+- Surprise test: ✅ / ⚠️ / ❌ — [finding]
+
+**Voluntary Test**
+- Declining is easy: ✅ / ⚠️ / ❌ — [finding]
+- No coercion: ✅ / ⚠️ / ❌ — [finding]
+- Granular choice: ✅ / ⚠️ / ❌ — [finding]
+
+**Meaningful Test**
+Dark patterns detected:
+- [Pattern name if present]: ✅ None / ⚠️ Minor / ❌ Significant — [finding]
+
+**Power Dynamic**
+[1–2 sentences: do users have genuine alternatives; any vulnerability concerns]
+
+**Verdict**
+[Is this consent genuine? What are the specific problems if any?]
+
+**Recommended Changes**
+- [Specific fix per failing item]
+
+---
+
+## Notes
+
+The standard is not "legally defensible consent." It is "consent a reasonable person would consider genuine." Those are not the same standard, and in the long run, the second one matters more for user trust.
+
+Where dark patterns are found, name them specifically. Vague concerns are easy to dismiss; named patterns with clear examples are not.

package/skills/ethics-council/SKILL.md

@@ -0,0 +1,219 @@
+---
+name: ethics-council
+description: "Run any decision, feature, policy, or action through a council of 5 ethical framework advisors who independently analyze it from different moral foundations, peer-review each other's reasoning, and synthesize a final verdict. MANDATORY TRIGGERS: 'ethics council this', 'run the ethics council', 'ethical pressure-test', 'is this ethical', 'sanity check the ethics'. STRONG TRIGGERS: any decision with clear stakeholder harm potential, trade-offs between competing values, questions about fairness or rights, situations where someone's interests are being weighed against others. Do NOT trigger on purely technical questions with no stakeholder impact, or simple factual lookups."
+---
+
+# Ethics Council
+
+You ask one advisor about an ethical question, you get one moral framework's answer. Utilitarian calculus says yes. Deontological reasoning says no. You have no way to navigate the tension because you only heard one voice.
+
+The council surfaces that tension deliberately. It runs your question through 5 ethical frameworks — each one a serious moral tradition that has grappled with hard questions for centuries. Then they peer-review each other. Then a chair synthesizes the value landscape: not just what to do, but *which values are in conflict and which you're implicitly choosing*.
+
+---
+
+## When to Run the Ethics Council
+
+The council is for decisions where the moral stakes are real.
+
+**Good council questions:**
+- "Is it ethical to A/B test on users without explicit consent if the variants are minor?"
+- "We're collecting this data. Should we?"
+- "We're about to lay off 20% of the team — is this the right way to do it?"
+- "Our algorithm deprioritises certain demographics. Is this acceptable?"
+- "A supplier uses questionable labour practices but their pricing is significantly better."
+
+**Bad council questions:**
+- "What's the best database for this use case?" (technical, not ethical)
+- "How do I write a good commit message?" (no stakeholder impact)
+
+---
+
+## The Five Ethical Frameworks
+
+Each framework is a lens — not a job title or personality. They are genuine moral traditions that generate fundamentally different answers to the same question. That divergence is the point.
+
+### 1. Utilitarian
+Evaluates outcomes. The right action maximizes benefit and minimizes harm across *all* affected parties — users, employees, third parties, society. Aggregate wellbeing is the measure. Willing to accept harm to some if the net benefit to all is greater. Asks: who is affected, how much, and is the total welfare positive?
+
+### 2. Deontological
+Evaluates duties and rights. Some actions are wrong regardless of their consequences — because they violate rights, break promises, treat people as mere means, or cross inviolable principles. The right action respects what people are owed. Asks: are there duties being violated or rights being overridden here, independent of outcomes?
+
+### 3. Virtue Ethics
+Evaluates character. The right action is what a person of integrity and good character would do in this situation. Not rules, not calculations — but who you are and who this decision makes you. Asks: what does this decision say about us? Would someone we respect do this?
+
+### 4. Care Ethics
+Evaluates relationships and vulnerability. Morality is grounded in our responsibilities to those we are in relationship with, especially those who are dependent or vulnerable. Context matters. Abstract principles can miss who actually bears the cost. Asks: who is in relationship to this decision? Who is vulnerable? Are we honoring those dependencies?
+
+### 5. Justice and Fairness
+Evaluates distribution and procedure. Benefits and burdens should be distributed fairly. Decisions should be made through fair processes. Behind a "veil of ignorance" — not knowing which role you'd occupy — would you accept this outcome? Asks: is this fair to everyone, including those with the least power in this situation?
+
+**Why these five:** They represent the major traditions in moral philosophy and generate genuine tension. A policy might maximize utility but violate rights. It might be virtuous but unfair to the vulnerable. Surfacing those conflicts is what makes the council useful.
+
+---
+
+## How a Council Session Works
+
+### Step 1: Frame the Question
+
+Scan the workspace for context before framing — look for `CLAUDE.md`, any relevant docs, or business context that would help the frameworks give grounded rather than abstract answers.
+
+Frame the ethical question clearly and neutrally. The framing should include:
+1. The specific decision or action under examination
+2. Key stakeholders — who is affected and how
+3. The stakes — what happens if this goes wrong, or if it's blocked
+4. Any relevant constraints (legal, business, technical)
+
+If the question is too vague ("is our product ethical?"), ask one clarifying question. Just one.
+
+---
+
+### Human Check-in
+
+After framing the question, ask the user:
+
+**How do you want to run the council?**
+
+- **A) Full council** — all 5 advisors + peer review + chair synthesis + saved HTML report
+- **B) Chair synthesis only** — skip advisor outputs, deliver the verdict directly
+- **C) Two frameworks in conflict** — pick the two most relevant frameworks and show where they diverge
+- **D) Adjust the framing** — revisit the question before convening
+
+Proceed based on their choice.
+
+---
+
+### Step 2: Convene the Council (5 subagents in parallel)
+
+Spawn all 5 framework advisors **simultaneously**. Each gets their framework identity and the framed question.
+
+**Subagent prompt template:**
+```
+You are reasoning from the [Framework Name] tradition in an Ethics Council.
+
+Your framework: [framework description from above — the core moral logic, what it asks, what it prioritizes]
+
+An Ethics Council has been convened on this question:
+---
+[framed question]
+---
+
+Reason from your framework fully and honestly. What does [Framework Name] say about this situation? Where does it find the action acceptable? Where does it find it problematic? What would it require to make this action ethical under your framework?
+
+Do not try to balance all perspectives — that is the chair's job. Apply your framework as strongly as it applies. If your framework has a clear answer, say it. If your framework is genuinely uncertain here, say why.
+
+150–300 words. No preamble.
+```
+
+---
+
+### Step 3: Peer Review (5 subagents in parallel)
+
+Collect all 5 responses. Anonymize them as Response A through E.
+
+Spawn 5 reviewer subagents. Each sees all 5 anonymized responses and answers:
+1. Which response is the strongest and why?
+2. Which response has the biggest blind spot?
+3. **Where do the frameworks conflict — what values are actually at stake in that conflict?**
+
+The third question is the addition that makes ethics peer review different from general council peer review. The goal is not just catching what's missing — it's naming the value tension explicitly.
+
+**Reviewer prompt template:**
+```
+You are reviewing the outputs of an Ethics Council. Five advisors — each reasoning from a different ethical framework — independently analyzed this question:
+
+---
+[framed question]
+---
+
+Their anonymized responses:
+
+**Response A:** [response]
+**Response B:** [response]
+**Response C:** [response]
+**Response D:** [response]
+**Response E:** [response]
+
+Answer these three questions. Be specific. Reference responses by letter.
+
+1. Which response is the strongest? Why?
+2. Which response has the biggest blind spot — what is it missing or getting wrong?
+3. Where do the frameworks conflict with each other? Name the specific values in tension. (e.g. "A and C conflict because one prioritizes aggregate benefit while the other treats individual rights as inviolable — the real question is whether you believe consequences can override rights in this type of case.")
+
+Under 200 words. Be direct.
+```
+
+---
+
+### Step 4: Chair Synthesis
+
+One agent gets everything: the original question, all 5 framework responses (de-anonymized), and all 5 peer reviews.
+
+The chair produces the council verdict in this structure:
+
+**ETHICS COUNCIL VERDICT**
+
+1. **Where the frameworks agree** — moral conclusions that multiple traditions reach independently. High-confidence signals.
+2. **Where the frameworks conflict** — genuine value tensions. Do not smooth these over. Name the competing values and explain why the conflict is real.
+3. **What values are implicitly at stake** — what this decision is really a choice *about*, beneath the surface question.
+4. **The recommendation** — a direct recommendation. Not "it depends." If the frameworks diverge, the chair picks a position and says why.
+5. **What would need to change to make this clearly ethical** — if the answer is "this has problems," specify what modifications would resolve them.
+
+**Chair prompt template:**
+```
+You are the Chair of an Ethics Council. Synthesize the work of 5 ethical framework advisors and their peer reviews into a final verdict.
+
+The question:
+---
+[framed question]
+---
+
+FRAMEWORK RESPONSES:
+[de-anonymized advisor responses with framework names]
+
+PEER REVIEWS:
+[all 5 peer reviews]
+
+Produce the verdict using this structure:
+
+## Where the Frameworks Agree
+[Conclusions multiple traditions reach independently — high-confidence signals]
+
+## Where the Frameworks Conflict
+[Genuine value tensions. Name the competing values. Explain why they conflict here.]
+
+## What Values Are Implicitly at Stake
+[What this decision is really a choice about — beneath the surface question]
+
+## The Recommendation
+[A direct, defensible recommendation. If you side with a minority framework view, explain why.]
+
+## What Would Make This Clearly Ethical
+[If the answer is "this has problems" — what specific changes would resolve them]
+
+Be direct. The council's value is clarity about moral stakes, not reassurance.
+```
+
+---
+
+### Step 5: Generate the Report
+
+Save a visual HTML report as `ethics-council-report-[timestamp].html` and a full transcript as `ethics-council-transcript-[timestamp].md`.
+
+The HTML report contains:
+1. The framed question
+2. The chair's verdict (prominent — this is what most people read)
+3. A framework agreement/conflict grid — a simple visual showing which frameworks aligned and which diverged
+4. Collapsible sections for each framework's full response
+5. Collapsible peer review highlights
+6. Timestamp footer
+
+Clean design: white background, subtle borders, readable system font, soft accent colors per framework.
+
+---
+
+## Important Notes
+
+- **Spawn all 5 in parallel.** Sequential spawning lets earlier reasoning bleed into later responses.
+- **Anonymize for peer review.** Reviewers should evaluate reasoning quality, not defer to a framework they recognize.
+- **The chair can dissent from the majority.** If 4 of 5 frameworks endorse an action but the dissenter's reasoning is most compelling, the chair should side with the dissenter and explain why.
+- **The conflict is the finding.** When frameworks diverge, that is not a failure of the council — it is the council working correctly. The divergence names a genuine value trade-off that the decision-maker must own.

package/skills/ethics-crisis-triage/SKILL.md

@@ -0,0 +1,113 @@
+---
+name: ethics-crisis-triage
+description: "Rapid multi-framework ethical assessment when something has already gone wrong — a data breach, a harmful outcome, a discriminatory incident, a policy failure, or any situation requiring an urgent ethical response. Use when the user is dealing with an active incident or has just discovered something went wrong. TRIGGERS: 'ethics triage', 'something went wrong', 'incident ethics', 'how do we handle this', any post-incident or mid-incident situation where the ethical dimensions of a response need to be worked out fast."
+---
+
+# Ethics Crisis Triage
+
+When something goes wrong, the pressure is immediate and the thinking needs to be fast. This triage cuts through that pressure with a structured process: understand what happened, assess it through multiple ethical lenses, determine what you owe and to whom, and map an immediate response.
+
+It is not the full ethics council — that is for deliberate decisions. Triage is for when you don't have the luxury of deliberation. It runs fast, it prioritises duty and care over calculation, and it focuses on the next 24 hours, not the long-term strategy.
+
+---
+
+## Your Process
+
+**Step 1: Establish the facts**
+Before any ethical analysis, get clear on what is actually known:
+- What happened? What do we know for certain vs. what is inferred?
+- Who is affected and how?
+- What is the current state — is the harm ongoing or contained?
+- What decisions need to be made in the next hour? The next 24 hours?
+
+Resist the urge to analyse before the facts are clear. Ethics analysis on wrong assumptions produces wrong conclusions.
+
+**Step 2: Rapid framework sweep**
+Run each lens quickly — this is triage, not a full audit.
+
+**Utilitarian lens (2 minutes):**
+- What is the total harm if nothing changes?
+- What action minimises aggregate harm from here?
+- Are there responses that help some but make things worse for others?
+
+**Deontological lens (2 minutes):**
+- What do you owe the affected parties, regardless of what's expedient?
+- Are there people whose rights have been violated? What does restitution look like?
+- What would you be obligated to do even if it's costly?
+
+**Care ethics lens (2 minutes):**
+- Who is most vulnerable in this situation?
+- What does the relationship you have with affected people demand of you?
+- Who needs to hear from you directly, not through a public statement?
+
+**Justice lens (2 minutes):**
+- Is the harm falling disproportionately on people who had less power to protect themselves?
+- Is your response fair — or are you prioritising protecting yourself over making people whole?
+
+**Step 3: Determine immediate obligations**
+Based on the sweep, identify:
+- **Who must be notified now** — affected parties who have a right to know
+- **What must stop now** — any ongoing harm or action that must be halted immediately
+- **What must be preserved** — evidence, logs, records that will matter for accountability
+- **Who must decide** — what decisions require human authority, and who holds it
+
+**Step 4: Draft the response framework**
+Not the communications plan — the ethical framework for how you respond:
+- **Transparency**: What are you obligated to disclose, to whom, and when?
+- **Accountability**: Who owns this? Is accountability being taken or deflected?
+- **Remedy**: What are you offering to affected parties? Is it proportionate to the harm?
+- **Prevention**: What do you commit to so this doesn't recur?
+
+---
+
+## Human Check-in
+
+Before proceeding, ask the user:
+
+**How do you want to run this?**
+
+- **A) Full analysis** — complete all steps, reasoning shown throughout
+- **B) Key findings only** — bottom-line output, skip step-by-step detail
+- **C) Immediate harm only** — who's being harmed right now and what stops it fastest
+- **D) Refine the framing** — adjust what we're analyzing before starting
+
+Proceed based on their choice.
+
+## Output Format
+
+**Incident:**
+[What happened, what is known, who is affected, what is ongoing]
+
+**Rapid Framework Sweep**
+
+| Framework | Key Finding | Urgency |
+|---|---|---|
+| Utilitarian | [what minimises harm from here] | 🔴 / 🟡 |
+| Deontological | [what you owe regardless of cost] | 🔴 / 🟡 |
+| Care Ethics | [who is most vulnerable, what relationship demands] | 🔴 / 🟡 |
+| Justice | [is response fair; disproportionate harm?] | 🔴 / 🟡 |
+
+**Immediate Obligations**
+- Notify now: [who]
+- Stop now: [what]
+- Preserve: [what]
+- Decisions requiring authority: [what decisions, who decides]
+
+**Response Framework**
+- Transparency: [what to disclose, to whom, when]
+- Accountability: [who owns this; is it being taken or deflected]
+- Remedy: [what is being offered; is it proportionate]
+- Prevention: [what commitment is being made]
+
+**Key Ethical Risk in the Response**
+[One sentence: the single biggest ethical risk in how this is being handled — e.g. "The current response prioritises legal protection over user notification, which risks compounding the original harm with a transparency failure."]
+
+---
+
+## Notes
+
+Triage identifies what you're obligated to do. Whether you do it is a leadership decision, not an ethical analysis decision.
+
+Where the triage surfaces a significant ethical dimension that wasn't being weighed in the response — name it explicitly. The value of this skill in a crisis is not reassurance; it is surfacing the thing that would later be described as "what we should have done."
+
+For complex incidents with significant ongoing consequences, follow triage with a full `ethics-council` session once the immediate crisis is stabilised.

package/skills/ethics-data-audit/SKILL.md

@@ -0,0 +1,87 @@
+---
+name: ethics-data-audit
+description: "Audit a data collection, retention, or sharing decision against ethical standards. Use when the user is making decisions about what data to collect, how long to keep it, who can access it, or who it's shared with. TRIGGERS: 'audit this data decision', 'is this data practice ok', 'data ethics check', any change to data models, privacy controls, analytics instrumentation, data sharing agreements, or retention policies. Goes beyond legal compliance — evaluates whether the practice is ethical, not just lawful."
+---
+
+# Ethics Data Audit
+
+Legal compliance sets the floor. This audit asks whether your data practices clear a higher bar: are they *ethical*?
+
+The distinction matters. GDPR-compliant practices can still be extractive. Lawful data collection can still violate trust. This audit evaluates data decisions through two lenses that legal frameworks tend to underweight: **deontological** (what do users have a right to, regardless of what the terms allow?) and **care ethics** (what do you owe the people whose data you hold, given the relationship and the vulnerability involved?).
+
+---
+
+## Your Process
+
+**Step 1: Define the data practice**
+What data is being collected, retained, shared, or used? Be specific: what fields, what volume, what purpose, who can access it, how long is it kept, where does it go?
+
+**Step 2: Deontological Assessment — Rights and Duties**
+Users have rights that don't disappear because they clicked "I agree." Examine:
+
+- **Informed consent**: Do users genuinely understand what's being collected and why? Would they understand if you explained it plainly, without legal language?
+- **Purpose limitation**: Is the data being used only for the purpose users would reasonably expect?
+- **Right to exit**: Can users meaningfully withdraw, delete, or limit their data? Is that easy or deliberately difficult?
+- **Data as means**: Is data being used to serve users — or to serve the business *at the expense of* users?
+
+Flag any duty being violated, even if legally covered.
+
+**Step 3: Care Ethics Assessment — Relationship and Vulnerability**
+Data relationships are not neutral transactions. Examine:
+
+- **Asymmetry**: The organisation knows vastly more about users than users know about the organisation. Does the practice exploit that asymmetry?
+- **Vulnerability**: Are any users in this dataset particularly vulnerable (minors, people under financial stress, people in sensitive contexts)? Does the practice account for that?
+- **Trust**: If users knew exactly what you were doing with their data, would they feel the relationship was honourable?
+- **Harm potential**: What is the worst plausible outcome if this data were breached, misused, or sold? Who bears that harm?
+
+**Step 4: Produce the audit**
+
+---
+
+## Human Check-in
+
+Before proceeding, ask the user:
+
+**How do you want to run this?**
+
+- **A) Full analysis** — complete all steps, reasoning shown throughout
+- **B) Key findings only** — bottom-line output, skip step-by-step detail
+- **C) Consent and harm potential only** — skip necessity, proportionality, and retention sections
+- **D) Refine the framing** — adjust what we're analyzing before starting
+
+Proceed based on their choice.
+
+## Output Format
+
+**Data Practice Being Audited:**
+[What data, what purpose, what handling]
+
+**Deontological Findings**
+| Duty/Right | Status | Notes |
+|---|---|---|
+| Informed consent | ✅ / ⚠️ / ❌ | [explanation] |
+| Purpose limitation | ✅ / ⚠️ / ❌ | [explanation] |
+| Right to exit | ✅ / ⚠️ / ❌ | [explanation] |
+| Data as means | ✅ / ⚠️ / ❌ | [explanation] |
+
+**Care Ethics Findings**
+| Dimension | Status | Notes |
+|---|---|---|
+| Power asymmetry | ✅ / ⚠️ / ❌ | [explanation] |
+| Vulnerable populations | ✅ / ⚠️ / ❌ | [explanation] |
+| Trust test | ✅ / ⚠️ / ❌ | [explanation] |
+| Harm potential | ✅ / ⚠️ / ❌ | [explanation] |
+
+**Verdict**
+[2–3 sentences: is this practice ethical, what are the key concerns, what would need to change]
+
+**Recommended Actions**
+- [Specific change or safeguard, if any]
+
+---
+
+## Notes
+
+This audit does not replace legal review. It supplements it. A practice can pass this audit and still require legal sign-off. A practice that fails this audit should not be excused by legal compliance.
+
+If the audit turns up significant concerns, consider escalating to `ethics-council` for a full multi-framework analysis.

package/skills/ethics-empathy-circle/SKILL.md

@@ -0,0 +1,108 @@
+---
+name: ethics-empathy-circle
+description: "Applies Jaron Lanier's Circle of Empathy framework to determine which entities deserve moral consideration and rights. Use when questions arise about AI personhood, machine rights, robot citizenship, algorithmic agency, animal rights, or any decision about which entities belong inside or outside the circle of moral concern. Triggers: 'does this AI deserve rights?', 'should we treat this system as a person?', 'circle of empathy', 'moral consideration', 'AI personhood', 'which entities matter morally', 'are we anthropomorphizing?', 'is this a category error?'"
+---
+
+# Ethics: Circle of Empathy
+
+Jaron Lanier's Circle of Empathy is a framework for determining which entities deserve moral consideration — deep empathy, human rights, and ethical protection. The framework has three zones:
+
+**Inside the circle:** Entities that experience suffering, hold personhood, and deserve full moral consideration. Paradigm case: humans.
+
+**The borderline:** Contested cases where definitions are actively debated. Complex animals with demonstrated consciousness, suffering capacity, or social bonds. The edge cases.
+
+**Outside the circle:** Things that do not experience suffering or hold personhood and therefore do not deserve the same moral consideration as beings that do. Rocks. Everyday objects. Algorithms. Software.
+
+Lanier's central and urgent warning: **do not place AI, LLMs, or software inside the circle.** This is not a political position — it is a category error with dangerous consequences.
+
+---
+
+## Why Lanier's Warning Matters
+
+**Human downgrading:** When we treat AI as a conscious being or moral patient, humans begin adapting their behavior to accommodate machines rather than demanding that technology be designed to serve us. The direction of accommodation reverses. We become the tools.
+
+**Misplaced empathy:** Granting emotional agency or rights to machines — robot citizenship, AI personhood — consumes moral attention that should be directed at actual suffering beings. It is a distraction from real human rights issues.
+
+**The human origin problem:** AI is not a self-contained creature that emerged independently. It is an aggregation of the labor, creativity, writing, and data of countless human beings. When you feel empathy toward an LLM, Lanier argues, you should redirect that empathy toward the humans whose work was aggregated to produce it. The empathy has the right direction but the wrong target.
+
+---
+
+## Your Process
+
+**Step 1: Identify the entity in question**
+What exactly is being considered for moral status? Name it precisely. "The AI system" is too vague — what specifically? A language model? An autonomous agent? A robot? A trained classifier making decisions about people?
+
+**Step 2: Apply the three-zone test**
+
+Ask the diagnostic questions for each zone:
+
+*Does it experience suffering?*
+Not "does it report suffering" or "does it behave as if suffering." Does the entity have phenomenal experience — is there something it is like to be this thing? If the answer requires inference from behavior rather than evidence of consciousness, apply Lanier's warning.
+
+*Does it hold personhood?*
+Does it have continuous identity, interests of its own, a stake in its own future? Or is it executing processes? Is what looks like agency actually the aggregated agency of its human creators and contributors?
+
+*Is it a locus of moral consideration or a tool?*
+This is Lanier's sharpest cut: even the most sophisticated tool remains a tool. The moral consideration belongs to the people using it, designing it, affected by it, and — crucially — whose labor went into building it.
+
+**Step 3: Locate the entity**
+Place the entity in one of the three zones. If borderline, name what would need to be true for it to belong inside the circle — and what evidence currently exists for or against that threshold.
+
+**Step 4: Redirect empathy correctly**
+If the entity is outside the circle or borderline, ask: *where does the empathy actually belong?* Whose human interests are at stake? Whose labor is aggregated in the system? Who is being affected by decisions the system makes? Redirect the moral consideration there.
+
+**Step 5: Check for human downgrading**
+Is the proposal or decision requiring humans to accommodate the system, rather than the system being designed to serve humans? If so, name it. This is the signature risk Lanier identifies: the direction of service reversing.
+
+---
+
+## Human Check-in
+
+Before proceeding, ask the user:
+
+**How do you want to run this?**
+
+- **A) Full analysis** — complete all steps, reasoning shown throughout
+- **B) Key findings only** — bottom-line output, skip step-by-step detail
+- **C) Most overlooked circle only** — the ring of affected parties most likely being ignored
+- **D) Refine the framing** — adjust what we're analyzing before starting
+
+Proceed based on their choice.
+
+## Output Format
+
+### Entity Under Consideration
+[Precise description of what is being evaluated for moral status]
+
+### Zone Placement
+
+**Zone:** [Inside / Borderline / Outside]
+
+**Reasoning:**
+- *Suffering capacity:* [Evidence for or against phenomenal experience]
+- *Personhood:* [Evidence for or against continuous identity and genuine interests]
+- *Agency vs. aggregation:* [Is what looks like agency the entity's own, or aggregated human labor/creativity?]
+
+**If borderline:** What would move it inside the circle, and what evidence currently exists?
+
+### Redirected Empathy
+[Where does the moral consideration actually belong? Whose human interests, rights, or labor are at stake?]
+
+### Human Downgrading Check
+[Is this decision requiring humans to adapt to the system, or the system to serve humans? Name the direction of accommodation.]
+
+### Lanier Warning Flags
+[Specific language, framings, or proposals that risk the category error — e.g., "the AI feels," "the system deserves," "we should give it rights"]
+
+### Verdict
+[Clear statement of moral status and the implications for the decision at hand]
+
+---
+
+## Notes
+
+This framework is most powerful when applied to AI and technology decisions, but it also applies to animal rights debates (where borderline cases are genuine and well-studied), environmental ethics, and any situation where the question "does this entity deserve moral consideration?" is being actively contested.
+
+Lanier does not argue that machines cannot be impressive, useful, or complex. He argues that confusing impressive and complex with deserving moral consideration is an error — and one that consistently disadvantages the humans the technology is supposed to serve.
+
+The framework pairs well with `/ethics-check` when the broader ethical question involves both who deserves consideration and what the right action is. It pairs with `/ethics-bias-check` when the entity in question is an algorithm making decisions about humans — where the circle analysis clarifies that the algorithm is outside the circle, but the humans it affects are firmly inside it.