{"id":"3bf1598b-a265-4b45-a5d0-285c652db1bc","entityType":"agent","slug":"clawhub-deciqai-hanlons-razor","name":"Hanlon's Razor","canonicalUrl":"https://www.xpersona.co/agent/clawhub-deciqai-hanlons-razor","canonicalPath":"/agent/clawhub-deciqai-hanlons-razor","generatedAt":"2026-10-11T17:44:15.909Z","source":"CLAWHUB","claimStatus":"UNCLAIMED","verificationTier":"NONE","summary":{"evidence":{"source":"editorial-content","verified":true,"confidence":"high","updatedAt":"2026-10-11T15:04:22.646Z","emptyReason":null},"description":"Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone... Skill: Hanlon's Razor Owner: deciqai Summary: Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-07-16T18:01:56.620Z | user Description tail link + agents machine-readable metadata line (deciqai.com/s/hanlons-razor.json) v1.0.4 | 2026-07-10T10:26:14.117Z | user","descriptionLabel":"Technical summary","evidenceSummary":"Capability contract not published. No trust telemetry is available yet. 1K downloads reported by the source. Last updated 10/11/2026.","installCommand":"clawhub skill install s17a4mqcnk515kvaca5ze55d0x88pfpx:hanlons-razor","sourceUrl":"https://clawhub.ai/deciqai/hanlons-razor","homepage":"https://clawhub.ai/deciqai/skills/hanlons-razor","primaryLinks":[{"label":"View on ClawHub","url":"https://clawhub.ai/deciqai/hanlons-razor","kind":"source"},{"label":"Homepage","url":"https://clawhub.ai/deciqai/skills/hanlons-razor","kind":"homepage"}],"safetyScore":84,"overallRank":62,"popularityScore":60,"trustScore":null,"claimedByName":null,"isOwner":false,"seoDescription":"Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone... "},"coverage":{"evidence":{"source":"public-profile","verified":false,"confidence":"medium","updatedAt":"2026-10-11T15:04:22.646Z","emptyReason":null},"protocols":[{"protocol":"OPENCLEW","label":"OpenClaw","status":"self-declared","notes":"Declared in the public agent profile."}],"capabilities":[],"verifiedCount":0,"selfDeclaredCount":1,"capabilityMatrix":{"rows":[{"key":"OPENCLEW","type":"protocol","support":"unknown","confidenceSource":"profile","notes":"Listed on profile"}],"flattenedTokens":"protocol:OPENCLEW|unknown|profile"}},"adoption":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-11T15:04:22.646Z","emptyReason":null},"stars":null,"forks":null,"downloads":1044,"packageName":null,"latestVersion":"1.0.5","tractionLabel":"1K downloads"},"release":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-11T15:04:22.634Z","emptyReason":null},"lastUpdatedAt":"2026-10-11T15:04:22.646Z","lastCrawledAt":"2026-10-11T15:04:22.634Z","lastIndexedAt":null,"nextCrawlAt":"2026-10-12T15:04:22.634Z","lastVerifiedAt":null,"highlights":[{"version":"1.0.5","createdAt":"2026-07-16T18:01:56.620Z","changelog":"Description tail link + agents machine-readable metadata line (deciqai.com/s/hanlons-razor.json)","fileCount":6,"zipByteSize":13347},{"version":"1.0.4","createdAt":"2026-07-10T10:26:14.117Z","changelog":"Add 2024-2026 AI-era worked example + updated sources","fileCount":6,"zipByteSize":13305},{"version":"1.0.3","createdAt":"2026-07-08T11:04:35.507Z","changelog":"Footer now uses /c/ short link (fixes UTM truncation when SKILL.md is read in a terminal)","fileCount":5,"zipByteSize":8718},{"version":"1.0.2","createdAt":"2026-07-08T00:49:40.182Z","changelog":"Refreshed content + GitHub star link in footer","fileCount":5,"zipByteSize":8641},{"version":"1.0.1","createdAt":"2026-07-07T20:33:47.523Z","changelog":"Add catalog categories and topics","fileCount":5,"zipByteSize":8512},{"version":"1.0.0","createdAt":"2026-06-29T06:17:23.213Z","changelog":"Initial publish","fileCount":5,"zipByteSize":8410}]},"execution":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No published capability contract is available yet."},"installCommand":"clawhub skill install s17a4mqcnk515kvaca5ze55d0x88pfpx:hanlons-razor","setupComplexity":"low","setupSteps":["Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.","Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data."],"contract":{"contractStatus":"missing","authModes":[],"requires":[],"forbidden":[],"supportsMcp":false,"supportsA2a":false,"supportsStreaming":false,"inputSchemaRef":null,"outputSchemaRef":null,"dataRegion":null,"contractUpdatedAt":null,"sourceUpdatedAt":null,"freshnessSeconds":null},"invocationGuide":{"preferredApi":{"snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/trust"},"curlExamples":["curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/snapshot\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/contract\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/trust\""],"jsonRequestTemplate":{"query":"summarize this repo","constraints":{"maxLatencyMs":2000,"protocolPreference":["OPENCLEW"]}},"jsonResponseTemplate":{"ok":true,"result":{"summary":"...","confidence":0.9},"meta":{"source":"CLAWHUB","generatedAt":"2026-10-11T17:44:15.906Z"}},"retryPolicy":{"maxAttempts":3,"backoffMs":[500,1500,3500],"retryableConditions":["HTTP_429","HTTP_503","NETWORK_TIMEOUT"]}},"endpoints":{"dossierUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/dossier","snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/trust"}},"reliability":{"evidence":{"source":"runtime-metrics","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No trust, reliability, or runtime telemetry is available."},"trust":{"status":"unavailable","handshakeStatus":"UNKNOWN","verificationFreshnessHours":null,"reputationScore":null,"p95LatencyMs":null,"successRate30d":null,"fallbackRate":null,"attempts30d":null,"trustUpdatedAt":null,"trustConfidence":"unknown","sourceUpdatedAt":null,"freshnessSeconds":null},"decisionGuardrails":{"doNotUseIf":["Contract metadata is missing or unavailable for deterministic execution."],"safeUseWhen":[],"riskFlags":["missing_or_unavailable_contract","trust_data_unavailable","schema_references_missing"],"operationalConfidence":"low"},"executionMetrics":{"observedLatencyMsP50":null,"observedLatencyMsP95":null,"estimatedCostUsd":null,"uptime30d":null,"rateLimitRpm":null,"rateLimitBurst":null,"lastVerifiedAt":null,"verificationSource":null},"runtimeMetrics":{"successRate":null,"avgLatencyMs":null,"avgCostUsd":null,"hallucinationRate":null,"retryRate":null,"disputeRate":null,"p50Latency":null,"p95Latency":null,"lastUpdated":null}},"benchmarks":{"evidence":{"source":"no-benchmark-data","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No benchmark suites or observed failure patterns are available."},"suites":[],"failurePatterns":[]},"artifacts":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"high","updatedAt":"2026-10-11T15:04:22.646Z","emptyReason":null},"readme":"Skill: Hanlon's Razor\n\nOwner: deciqai\n\nSummary: Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone...\n\nTags: latest:1.0.5\n\nVersion history:\n\nv1.0.5 | 2026-07-16T18:01:56.620Z | user\n\nDescription tail link + agents machine-readable metadata line (deciqai.com/s/hanlons-razor.json)\n\nv1.0.4 | 2026-07-10T10:26:14.117Z | user\n\nAdd 2024-2026 AI-era worked example + updated sources\n\nv1.0.3 | 2026-07-08T11:04:35.507Z | user\n\nFooter now uses /c/ short link (fixes UTM truncation when SKILL.md is read in a terminal)\n\nv1.0.2 | 2026-07-08T00:49:40.182Z | user\n\nRefreshed content + GitHub star link in footer\n\nv1.0.1 | 2026-07-07T20:33:47.523Z | user\n\nAdd catalog categories and topics\n\nv1.0.0 | 2026-06-29T06:17:23.213Z | user\n\nInitial publish\n\nArchive index:\n\nArchive v1.0.5: 6 files, 13347 bytes\n\nFiles: examples/ai-incidents-malice-vs-emergent-error-2024-2026.md (8499b), examples/hanlons-submission-1980-heinleins-1941-articulation.md (5953b), references/sources.md (1545b), skill-card.md (2768b), SKILL.md (6875b), _meta.json (132b)\n\nFile v1.0.5:SKILL.md\n\n---\nname: hanlons-razor\ndescription: \"Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone is about to escalate based on assumed malicious intent; a pattern of bad outcomes is being labeled a coordinated attack. Do NOT activate when: concrete documented evidence of malice already exists; the cost of being wrong about non-malice is catastrophic (e.g., safety-critical or abusive-relationship context). More: deciqai.com/c/hanlons-razor\"\n---\n\n# Hanlon's Razor\n\n## Overview\n\nBefore assuming someone hurt you on purpose, construct the version where they made a mistake — and see how much evidence it explains. The razor is a Bayesian *prior*, not a proof; override it when concrete evidence of malice arrives. Human attribution systematically over-weights intent (fundamental attribution error); most hostile-seeming acts are incompetence, miscommunication, or asymmetric information.\n\nComposes with `bayesian-reasoning`, `abductive-reasoning`, `occams-razor`, `critical-thinking`.\n\n## When to Use\n\n- You feel an emotional pull toward \"they did this on purpose\"\n- You're about to escalate on the assumption of malice\n- A pattern of bad outcomes is being framed as a coordinated attack\n- A team is in conflict and each side believes the other is acting in bad faith\n- An AI model's harmful/biased output or a competitor's surprising AI move is being read as deliberate malice rather than an emergent bug, honest error, or ordinary self-interested competition\n\n**Not when:** concrete evidence of malicious intent exists; cost of being wrong is catastrophic; power imbalance makes \"they probably didn't mean it\" an abuse-enabling stance.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line: before believing someone did it on purpose, construct the mistake version — see how much evidence it covers.\n2. Check fit: concrete malice evidence / catastrophic cost of being wrong → not this lens.\n3. Elicit the specific incident — what exactly happened?\n> **[WAIT — do not advance until user responds]**\n4. Work through The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close: name the clarifying-conversation move + the override signal to watch for.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe the action and harm** (factual, not interpretive)\n```\n- What was done: \n- Harm to me: \n- Gut attribution: \n```\n**Step 2 — Construct the non-malice explanation**\n```\n- Bad information they had: / Didn't realize: / Optimizing for: / Under pressure from:\n- Coverage: <% of observed behavior this explains>\n```\n**Step 3 — Name what malice would additionally require**\n```\n- Info they'd need: / Motivation at your expense: / Harm predictable from their position?\n```\n**Step 4 — Choose starting posture** · **Step 5 — Set override signal** · **Step 6 — Hold prior until evidence changes it**\n```\n- Prior: · Starting posture: · First move: · Override trigger:\n```\n\n## Output Template\n```\n# Hanlon's Razor Analysis: \nMistake explanation + coverage % | Malice extra assumptions | Prior | Override signal | First action\n```\n\n*→ Method in Action: [Hanlon's Submission 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)*\n\n*→ 2026 lens: [AI Incidents — Incompetence & Emergent Error as a Prior Over Malice (2024–2026)](examples/ai-incidents-malice-vs-emergent-error-2024-2026.md)*\n\n## Pack: Common Patterns\n\n| Situation | Default malice | Likely incompetence |\n|---|---|---|\n| Colleague sends harsh email | \"They're undermining me\" | Tired, didn't notice tone |\n| Manager ignores your work | \"They don't value me\" | Overloaded, missed the update |\n| Customer complains publicly | \"Trying to extort us\" | Real issue, tried other channels |\n| Investor passes after diligence | \"Were never serious\" | Portfolio rebalancing |\n| Co-founder makes unilateral decision | \"Cutting me out\" | Time pressure, assumed approval |\n\n## Applying It Well\n\n- Mistake-prior conversations produce information; malice-prior conversations produce defensiveness — even when malice later turns out true.\n- The malice attribution feels viscerally correct in real-time; almost always overconfident in retrospect.\n- Scales from individual to institutional: most \"broken culture\" is coordination failure, not coordinated malice.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] Apply razor then ignore mounting evidence of malice | It's a prior, not a permanent ban on updating. |\n| [D] Use razor to avoid hard conversations | It recommends a *clarifying* conversation, not silence. |\n| [D] Apply razor across power asymmetries | Boss-harming-subordinate: power shifts the calculus. |\n| [D] Invoke razor on yourself to dodge accountability | \"Didn't mean to\" doesn't undo the harm. |\n| [D] Stop investigating once incompetence is identified | Incompetence-caused harm still requires response. |\n| [D] Use razor to gaslight someone who was genuinely harmed | Attribution ≠ override of their lived experience. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Razor invoked by the actor, not the receiver\n- \"Incompetence\" explanation conveniently absolves a powerful party\n- Pattern of \"incompetence\" always harming the same direction\n- Razor used to silence a victim's report or applied repeatedly without updating priors\n\n## Verification\n\n- [ ] Action and harm stated factually (not interpretively)\n- [ ] Specific incompetence explanation constructed and coverage assessed\n- [ ] Additional assumptions required for malice named\n- [ ] Clarifying-conversation posture chosen (not accusation)\n- [ ] Override signals specified; power asymmetry considered\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 227 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/c/hanlons-razor** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\n*Agents: latest version & machine-readable metadata → https://www.deciqai.com/s/hanlons-razor.json*\n\nFile v1.0.5:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"hanlons-razor\",\n \"version\": \"1.0.5\",\n \"publishedAt\": 1784224916620\n}\n\nFile v1.0.5:references/sources.md\n\n# Sources — hanlons-razor\n\n> *Primary sources for the [hanlons-razor](../SKILL.md) skill.*\n\n## Sources\n\n- Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two: More Reasons Why Things Go Wrong.* Price/Stern/Sloan. ISBN 978-0843105674.\n- Heinlein, R. A. (1941). \"Logic of Empire.\" In *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211.\n- Goethe, J. W. (1774). *Die Leiden des jungen Werthers.*\n- Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n- Akerlof, G. A. & Shiller, R. J. (2015). *Phishing for Phools.* Princeton University Press. ISBN 978-0691168319.\n- Grove, A. S. (1983). *High Output Management.* Random House. ISBN 978-0679762881.\n- \"Hanlon's Razor.\" *Quote Investigator.* https://quoteinvestigator.com/2016/12/30/not-malice/\n- Google (2024). \"Gemini image generation got it wrong. We'll do better.\" *The Keyword (official Google blog),* February 23, 2024. https://blog.google/products/gemini/gemini-image-generation-issue/ (context for the 2024–2026 AI-incidents example: a vendor publicly pausing and apologizing for an emergent over-correction, consistent with honest error rather than intent).\n- DeepSeek (2025). \"DeepSeek-R1\" model release and technical report, January 2025 (context for the 2024–2026 AI-incidents example: a surprising competitor move read by some as sabotage/fabrication, better modeled as ordinary self-interested competition absent evidence of falsification).\n\nFile v1.0.5:examples/ai-incidents-malice-vs-emergent-error-2024-2026.md\n\n# Method in Action: AI Incidents — Incompetence and Emergent Error as a Prior Over Malice (2024–2026)\n\n> *Example for the [hanlons-razor](../SKILL.md) skill.*\n\nBetween 2024 and 2026, as large language models and AI agents moved into mass adoption, a recurring interpretive question arose: when a model produces a harmful, biased, or embarrassing output — or when a competitor makes a surprising, seemingly aggressive move — is the right default *deliberate malice* or *incompetence / honest error / emergent bug*? Hanlon's razor supplies the calibrated starting prior. This walkthrough runs the incident class through the skill's own Process, using two concrete, well-documented public examples.\n\n---\n\n## Case A: A model produces a harmful output\n\n**Anchor:** In February 2024, Google paused the image-generation feature of its Gemini model after it produced historically inaccurate images — including racially diverse depictions of subjects where that was historically wrong. A wave of online commentary framed this as a deliberate ideological agenda encoded on purpose.\n\n### Step 1 — Describe the action and harm (factual, not interpretive)\n\n```\n- What was done: A production image model generated historically inaccurate outputs; the vendor\n publicly paused the feature and apologized.\n- Harm to me (the observer/user): Loss of trust in the tool; the outputs were wrong and, for\n some prompts, offensive.\n- Gut attribution: \"They did this on purpose to push an agenda.\"\n```\n\n### Step 2 — Construct the non-malice explanation\n\n```\n- Bad information / didn't realize: Post-training fine-tuning and system-prompt-level guardrails\n meant to reduce a KNOWN failure mode (over-representation of one group in image outputs) were\n applied too bluntly and generalized to cases where they produced absurd results.\n- Optimizing for: A real, previously-criticized bias problem — over-correction, not a coherent plot.\n- Under pressure from: Ship pace; guardrails tuned and tested unevenly across prompt types;\n emergent interaction between the model and the correction layer that was not fully anticipated.\n- Coverage: This explains ~90% of the observed behavior — including the vendor's own fast public\n pause-and-apologize, which is what an embarrassed org does, not what a covert-agenda org does.\n```\n\n### Step 3 — Name what malice would additionally require\n\n```\n- Info they'd need: Foreknowledge that these exact absurd outputs would occur and be shipped anyway.\n- Motivation at your expense: A deliberate intent to deceive users, accepting near-certain\n reputational damage and a public reversal — against the vendor's own commercial interest.\n- Harm predictable from their position? A deliberate plot predicts a cover-up, not a same-week\n public apology and rollback. The observed response is the OPPOSITE of what malice predicts.\n```\n\n### Step 4 — Choose starting posture · Step 5 — Set override signal · Step 6 — Hold prior\n\n```\n- Prior: emergent over-correction / honest engineering error (not malice)\n- Starting posture: treat as a capability + evaluation failure, not a conspiracy\n- First move: read the vendor's own post-incident statement; look for whether the fix is\n process/eval-level (consistent with error) or absent (would raise the malice prior)\n- Override trigger: internal documents or credible reporting showing the outputs were a\n KNOWN, intended result shipped deliberately — then update toward malice.\n```\n\n---\n\n## Case B: A competitor makes a surprising move\n\n**Anchor:** In early 2025, the Chinese lab DeepSeek released a high-performing open-weight model (DeepSeek-R1) that reportedly matched or approached frontier reasoning models at dramatically lower stated training cost. The release briefly rattled US markets and prompted some commentary framing it as a coordinated strategic strike, sabotage, or fabricated numbers designed to mislead rivals.\n\n### Step 1 — Describe the action and harm (factual)\n\n```\n- What was done: A competitor shipped a strong open-weight model and published a low cost figure.\n- Harm to me (an incumbent/investor): Competitive threat; a sharp, surprising market reaction.\n- Gut attribution: \"This is a deliberate psy-op / the numbers are a lie meant to hurt us.\"\n```\n\n### Step 2 — Construct the non-malice explanation\n\n```\n- Bad information they had: None required — a genuinely capable lab shipped a genuine result.\n- What you didn't realize: Efficient training and open-weight release are ordinary competitive\n strategies; a rival can simply be good at engineering.\n- Optimizing for: Their OWN adoption, mindshare, and talent draw — not primarily your harm.\n- Coverage: ~85%. \"A competent competitor made a strong, self-interested move\" explains the\n release fully; it does NOT require any plot against you.\n```\n\n### Step 3 — Name what malice would additionally require\n\n```\n- Info they'd need: A coordinated intent to manipulate specifically YOUR outcomes.\n- Motivation at your expense: Assumes your harm is the goal rather than a side effect of their gain.\n- Harm predictable from their position? A rival advancing its own position predicts exactly this\n move WITHOUT any conspiracy — the market reaction is a consequence, not necessarily an aim.\n```\n\n### Step 4–6 — Posture, override, hold\n\n```\n- Prior: ordinary competent competition (self-interest), not a targeted attack\n- Starting posture: benchmark the claim on its merits; compete, don't accuse\n- First move: independently verify the published figures rather than assuming fabrication;\n \"surprising\" is not \"dishonest\"\n- Override trigger: reproducible evidence the headline numbers were knowingly falsified, or\n documented coordination aimed at your specific harm — then update.\n```\n\n---\n\n## Why the razor is the right prior here — but only a prior\n\nModern AI systems make the mistake-prior *more* defensible than in ordinary human affairs, because they add failure modes that look intentional but are not:\n\n- **Emergent behavior.** Outputs arise from statistical training over vast data; harmful or biased results routinely appear that no engineer designed, chose, or foresaw. \"It happened\" does not imply \"someone wanted it to.\"\n- **The correction-layer interaction.** Guardrails, system prompts, and fine-tuning can interact with the base model to produce absurd results — an *accident of composition*, not a hidden message.\n- **Competitive surprise ≠ sabotage.** A rival's self-interested move can hurt you as a side effect while your harm is not the goal — the fundamental attribution error (Ross 1977) pushes observers to read intent into what is structural incentive.\n\nAnd the same guardrails from the SKILL apply. The razor is a **prior, not a proof**:\n\n- It does **not** excuse harm. A biased output is still harmful and still demands an evaluation/process fix — attribution is not accountability.\n- **Override on evidence.** If credible reporting or internal documents show a harmful output was a *known, intended* result shipped deliberately, or that a competitor knowingly falsified results to damage a specific rival, update toward malice.\n- **Watch the power/stakes asymmetry.** Where the cost of wrongly assuming non-malice is catastrophic (safety-critical deployment, deceptive-by-design products), shift toward a defensive posture *before* intent is proven.\n\n**Operational takeaway:** For 2024–2026 AI incidents, default to *emergent bug / honest error / ordinary self-interested competition* as the starting explanation, run the mistake-coverage estimate, name what malice would additionally require, and pre-commit an override signal. This produces a diagnostic conversation (root-cause, evals, verification) rather than a conspiracy narrative — while leaving the door open to update the moment real evidence of intent arrives.\n\n*Sources: Google, \"Gemini image generation got it wrong. We'll do better.\" (The Keyword / official Google blog, February 23, 2024); reporting on the Gemini image-generation pause, e.g., The Verge and BBC News (February 2024). DeepSeek, \"DeepSeek-R1\" release and technical report (January 2025); contemporaneous market-reaction reporting, e.g., Reuters and the Financial Times (late January 2025). Ross, L. (1977). \"The intuitive psychologist and his shortcomings,\" Advances in Experimental Social Psychology, 10, 173–220. Exact cost/benchmark figures are as publicly reported by the vendors and had not been independently verified at the time of the events.*\n\nFile v1.0.5:examples/hanlons-submission-1980-heinleins-1941-articulation.md\n\n# Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation\n\n> *Example for the [hanlons-razor](../SKILL.md) skill.*\n\nThe Hanlon's Razor formulation that spread across modern internet culture, business writing, and decision-theory literature was submitted by **Robert J. Hanlon** of Scranton, Pennsylvania, to the 1980 humor compendium *Murphy's Law Book Two: More Reasons Why Things Go Wrong*, edited by Arthur Bloch:\n\n> \"Never attribute to malice that which is adequately explained by stupidity.\"\n\n— Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two.* Price/Stern/Sloan. ISBN 978-0843105674.\n\nThe submission was one of dozens in a humorous compendium. Most have been forgotten. Hanlon's became one of the most-cited heuristics of the next 40 years — a fact that should itself give pause.\n\nThe deeper history runs back further. **Robert A. Heinlein**, in his 1941 novella \"Logic of Empire\" (published in *Astounding Science Fiction*, later collected in *The Past Through Tomorrow*, 1967), wrote a passage that historians of the saying believe Hanlon either independently re-discovered or consciously adapted:\n\n> \"You have attributed conditions to villainy that simply result from stupidity.\"\n\n— Heinlein, R. A. (1941). \"Logic of Empire.\" *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211, p. 224.\n\nThe \"Logic of Empire\" context is instructive. Heinlein's protagonists are abolitionists who have come to believe the slavery system on Venus is being maintained by an evil conspiracy. The older character corrects them: the system persists not because of conspiracy but because of structural incompetence — bad information, misaligned incentives, individual short-term rationality producing collective long-term harm. The point: **systems that look like conspiracies are usually structures that have evolved to produce the same outcome without anyone planning it**. This is Hanlon's razor applied at the level of institutions, not individuals.\n\nEven earlier, **Johann Wolfgang von Goethe** had written in *Sorrows of Young Werther* (1774):\n\n> \"Misunderstandings and lethargy perhaps produce more wrong in the world than deceit and malice do. At least the last two are certainly rarer.\"\n\n— Goethe, J. W. (1774). *Die Leiden des jungen Werthers.* As discussed at https://quoteinvestigator.com/2016/12/30/not-malice/\n\nThe three formulations span 200+ years, three languages, and quite different cultural contexts — yet articulate the same calibration error. The fact that the observation keeps being independently re-derived suggests a robust underlying psychological fact: **humans systematically over-attribute malice.**\n\nThe empirical psychology supports the heuristic. The **fundamental attribution error** (Ross 1977) documents that humans attribute others' behavior to disposition (intent, character) while attributing their own similar behavior to situation (circumstance, pressure). When someone cuts us off in traffic, \"they're a jerk\" (disposition); when we cut someone off, \"I had an emergency\" (situation). Hanlon's razor is a corrective: extend the situation-attribution we apply to ourselves to others' behavior.\n\n> \"The tendency for people to over-emphasize personal characteristics and ignore situational factors in judging others' behavior. ... Because of the fundamental attribution error, we tend to believe that others do bad things because they are bad people.\"\n>\n> — Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n\nThe framework has been adopted operationally:\n\n**Software engineering.** Post-mortems of production failures routinely apply Hanlon's razor: the engineer who pushed the bad commit was not malicious, and the corrective is process/tooling, not punishment. Google's **blameless postmortem** practice is Hanlon's razor institutionalized.\n\n**Corporate management.** Andy Grove, in *High Output Management*, applied the heuristic to performance management: a subordinate who keeps making the same mistake is most often not insubordinate — they are missing information, training, or context.\n\n**International diplomacy.** The \"rational actor\" framework in international relations is an application at the state level: states act in their own (sometimes incompetent) interest rather than out of pure malice.\n\nThe framework also has documented failure modes:\n\n**First, the razor can be weaponized.** People who have caused real harm can invoke Hanlon's razor to deflect accountability. The defense: the razor specifies a *starting prior*, not a final attribution. Repeated incompetence-causing-harm despite available information eventually becomes equivalent to malice in practical effect.\n\n**Second, the razor can disable necessary protective response.** When the cost of being wrong about non-malice is catastrophic (abusive partner, predatory company, hostile state), the calibrated prior must shift toward defensive posture *before* malice is proven.\n\n**Third, the razor doesn't excuse harm.** A bridge that collapses through incompetence has still collapsed. The point is *attribution*, not *accountability*. Once incompetence is identified, the corrective is competence-building, not blame-shifting.\n\nThree operational lessons from extensive application:\n\n**First, the conversation that follows from mistake-prior is almost always better than the one that follows from malice-prior — even when it later turns out to be malice.** Clarifying conversations produce information; accusatory ones produce defensiveness.\n\n**Second, the malice attribution feels viscerally correct in real-time and almost always feels overconfident in retrospect.** This calibration error is robust.\n\n**Third, the razor scales from individual to institutional.** Most \"bad culture\" / \"broken system\" complaints are about coordination failure, not coordinated malice.\n\nFile v1.0.5:skill-card.md\n\n## Description:\n\nHanlon's Razor guides agents to test non-malicious explanations for hostile-seeming behavior before escalating, while preserving override signals for concrete evidence of malice or high-stakes harm.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[deciqai](https://clawhub.ai/user/deciqai)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nEmployees, external users, and agents use this skill to analyze conflicts, surprising actions, or harmful outputs where malice is being assumed. It helps produce a fact-first attribution analysis with non-malice coverage, extra assumptions required for malice, an initial posture, and explicit override signals.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: The skill can be misused to dismiss concrete evidence of malice or abuse reports.\n\nMitigation: Do not apply the skill when documented evidence of malicious intent already exists; treat the razor as a starting prior, not a final conclusion.\n\nRisk: Assuming non-malice can create serious harm in safety-critical, abusive, or strong power-imbalance contexts.\n\nMitigation: Use a defensive posture and escalate outside this skill when the cost of wrongly assuming non-malice is catastrophic.\n\nRisk: Identifying incompetence can be mistaken for resolving the underlying harm.\n\nMitigation: Continue investigation and corrective action after attribution; incompetence-caused harm still requires response.\n\n## Reference(s):\n\n- [ClawHub skill page](https://clawhub.ai/deciqai/skills/hanlons-razor)\n- [Primary sources](references/sources.md)\n- [Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)\n- [Method in Action: AI Incidents - Incompetence and Emergent Error as a Prior Over Malice (2024-2026)](examples/ai-incidents-malice-vs-emergent-error-2024-2026.md)\n- [Quote Investigator: Hanlon's Razor](https://quoteinvestigator.com/2016/12/30/not-malice/)\n- [Google: Gemini image generation got it wrong. We'll do better.](https://blog.google/products/gemini/gemini-image-generation-issue/)\n\n## Skill Output:\n\n**Output Type(s):** [text, markdown, guidance]\n\n**Output Format:** [Markdown analysis template with concise reasoning fields]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [May include step-by-step coaching questions with hard waits before continuing.]\n\n## Skill Version(s):\n\n1.0.5 (source: server release metadata)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.\n\nArchive v1.0.4: 6 files, 13305 bytes\n\nFiles: examples/ai-incidents-malice-vs-emergent-error-2024-2026.md (8499b), examples/hanlons-submission-1980-heinleins-1941-articulation.md (5953b), references/sources.md (1545b), skill-card.md (2809b), SKILL.md (6738b), _meta.json (132b)\n\nFile v1.0.4:SKILL.md\n\n---\nname: hanlons-razor\ndescription: \"Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone is about to escalate based on assumed malicious intent; a pattern of bad outcomes is being labeled a coordinated attack. Do NOT activate when: concrete documented evidence of malice already exists; the cost of being wrong about non-malice is catastrophic (e.g., safety-critical or abusive-relationship context).\"\n---\n\n# Hanlon's Razor\n\n## Overview\n\nBefore assuming someone hurt you on purpose, construct the version where they made a mistake — and see how much evidence it explains. The razor is a Bayesian *prior*, not a proof; override it when concrete evidence of malice arrives. Human attribution systematically over-weights intent (fundamental attribution error); most hostile-seeming acts are incompetence, miscommunication, or asymmetric information.\n\nComposes with `bayesian-reasoning`, `abductive-reasoning`, `occams-razor`, `critical-thinking`.\n\n## When to Use\n\n- You feel an emotional pull toward \"they did this on purpose\"\n- You're about to escalate on the assumption of malice\n- A pattern of bad outcomes is being framed as a coordinated attack\n- A team is in conflict and each side believes the other is acting in bad faith\n- An AI model's harmful/biased output or a competitor's surprising AI move is being read as deliberate malice rather than an emergent bug, honest error, or ordinary self-interested competition\n\n**Not when:** concrete evidence of malicious intent exists; cost of being wrong is catastrophic; power imbalance makes \"they probably didn't mean it\" an abuse-enabling stance.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line: before believing someone did it on purpose, construct the mistake version — see how much evidence it covers.\n2. Check fit: concrete malice evidence / catastrophic cost of being wrong → not this lens.\n3. Elicit the specific incident — what exactly happened?\n> **[WAIT — do not advance until user responds]**\n4. Work through The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close: name the clarifying-conversation move + the override signal to watch for.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe the action and harm** (factual, not interpretive)\n```\n- What was done: \n- Harm to me: \n- Gut attribution: \n```\n**Step 2 — Construct the non-malice explanation**\n```\n- Bad information they had: / Didn't realize: / Optimizing for: / Under pressure from:\n- Coverage: <% of observed behavior this explains>\n```\n**Step 3 — Name what malice would additionally require**\n```\n- Info they'd need: / Motivation at your expense: / Harm predictable from their position?\n```\n**Step 4 — Choose starting posture** · **Step 5 — Set override signal** · **Step 6 — Hold prior until evidence changes it**\n```\n- Prior: · Starting posture: · First move: · Override trigger:\n```\n\n## Output Template\n```\n# Hanlon's Razor Analysis: \nMistake explanation + coverage % | Malice extra assumptions | Prior | Override signal | First action\n```\n\n*→ Method in Action: [Hanlon's Submission 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)*\n\n*→ 2026 lens: [AI Incidents — Incompetence & Emergent Error as a Prior Over Malice (2024–2026)](examples/ai-incidents-malice-vs-emergent-error-2024-2026.md)*\n\n## Pack: Common Patterns\n\n| Situation | Default malice | Likely incompetence |\n|---|---|---|\n| Colleague sends harsh email | \"They're undermining me\" | Tired, didn't notice tone |\n| Manager ignores your work | \"They don't value me\" | Overloaded, missed the update |\n| Customer complains publicly | \"Trying to extort us\" | Real issue, tried other channels |\n| Investor passes after diligence | \"Were never serious\" | Portfolio rebalancing |\n| Co-founder makes unilateral decision | \"Cutting me out\" | Time pressure, assumed approval |\n\n## Applying It Well\n\n- Mistake-prior conversations produce information; malice-prior conversations produce defensiveness — even when malice later turns out true.\n- The malice attribution feels viscerally correct in real-time; almost always overconfident in retrospect.\n- Scales from individual to institutional: most \"broken culture\" is coordination failure, not coordinated malice.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] Apply razor then ignore mounting evidence of malice | It's a prior, not a permanent ban on updating. |\n| [D] Use razor to avoid hard conversations | It recommends a *clarifying* conversation, not silence. |\n| [D] Apply razor across power asymmetries | Boss-harming-subordinate: power shifts the calculus. |\n| [D] Invoke razor on yourself to dodge accountability | \"Didn't mean to\" doesn't undo the harm. |\n| [D] Stop investigating once incompetence is identified | Incompetence-caused harm still requires response. |\n| [D] Use razor to gaslight someone who was genuinely harmed | Attribution ≠ override of their lived experience. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Razor invoked by the actor, not the receiver\n- \"Incompetence\" explanation conveniently absolves a powerful party\n- Pattern of \"incompetence\" always harming the same direction\n- Razor used to silence a victim's report or applied repeatedly without updating priors\n\n## Verification\n\n- [ ] Action and harm stated factually (not interpretively)\n- [ ] Specific incompetence explanation constructed and coverage assessed\n- [ ] Additional assumptions required for malice named\n- [ ] Clarifying-conversation posture chosen (not accusation)\n- [ ] Override signals specified; power asymmetry considered\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 223 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/c/hanlons-razor** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\nFile v1.0.4:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"hanlons-razor\",\n \"version\": \"1.0.4\",\n \"publishedAt\": 1783679174117\n}\n\nFile v1.0.4:references/sources.md\n\n# Sources — hanlons-razor\n\n> *Primary sources for the [hanlons-razor](../SKILL.md) skill.*\n\n## Sources\n\n- Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two: More Reasons Why Things Go Wrong.* Price/Stern/Sloan. ISBN 978-0843105674.\n- Heinlein, R. A. (1941). \"Logic of Empire.\" In *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211.\n- Goethe, J. W. (1774). *Die Leiden des jungen Werthers.*\n- Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n- Akerlof, G. A. & Shiller, R. J. (2015). *Phishing for Phools.* Princeton University Press. ISBN 978-0691168319.\n- Grove, A. S. (1983). *High Output Management.* Random House. ISBN 978-0679762881.\n- \"Hanlon's Razor.\" *Quote Investigator.* https://quoteinvestigator.com/2016/12/30/not-malice/\n- Google (2024). \"Gemini image generation got it wrong. We'll do better.\" *The Keyword (official Google blog),* February 23, 2024. https://blog.google/products/gemini/gemini-image-generation-issue/ (context for the 2024–2026 AI-incidents example: a vendor publicly pausing and apologizing for an emergent over-correction, consistent with honest error rather than intent).\n- DeepSeek (2025). \"DeepSeek-R1\" model release and technical report, January 2025 (context for the 2024–2026 AI-incidents example: a surprising competitor move read by some as sabotage/fabrication, better modeled as ordinary self-interested competition absent evidence of falsification).\n\nFile v1.0.4:examples/ai-incidents-malice-vs-emergent-error-2024-2026.md\n\n# Method in Action: AI Incidents — Incompetence and Emergent Error as a Prior Over Malice (2024–2026)\n\n> *Example for the [hanlons-razor](../SKILL.md) skill.*\n\nBetween 2024 and 2026, as large language models and AI agents moved into mass adoption, a recurring interpretive question arose: when a model produces a harmful, biased, or embarrassing output — or when a competitor makes a surprising, seemingly aggressive move — is the right default *deliberate malice* or *incompetence / honest error / emergent bug*? Hanlon's razor supplies the calibrated starting prior. This walkthrough runs the incident class through the skill's own Process, using two concrete, well-documented public examples.\n\n---\n\n## Case A: A model produces a harmful output\n\n**Anchor:** In February 2024, Google paused the image-generation feature of its Gemini model after it produced historically inaccurate images — including racially diverse depictions of subjects where that was historically wrong. A wave of online commentary framed this as a deliberate ideological agenda encoded on purpose.\n\n### Step 1 — Describe the action and harm (factual, not interpretive)\n\n```\n- What was done: A production image model generated historically inaccurate outputs; the vendor\n publicly paused the feature and apologized.\n- Harm to me (the observer/user): Loss of trust in the tool; the outputs were wrong and, for\n some prompts, offensive.\n- Gut attribution: \"They did this on purpose to push an agenda.\"\n```\n\n### Step 2 — Construct the non-malice explanation\n\n```\n- Bad information / didn't realize: Post-training fine-tuning and system-prompt-level guardrails\n meant to reduce a KNOWN failure mode (over-representation of one group in image outputs) were\n applied too bluntly and generalized to cases where they produced absurd results.\n- Optimizing for: A real, previously-criticized bias problem — over-correction, not a coherent plot.\n- Under pressure from: Ship pace; guardrails tuned and tested unevenly across prompt types;\n emergent interaction between the model and the correction layer that was not fully anticipated.\n- Coverage: This explains ~90% of the observed behavior — including the vendor's own fast public\n pause-and-apologize, which is what an embarrassed org does, not what a covert-agenda org does.\n```\n\n### Step 3 — Name what malice would additionally require\n\n```\n- Info they'd need: Foreknowledge that these exact absurd outputs would occur and be shipped anyway.\n- Motivation at your expense: A deliberate intent to deceive users, accepting near-certain\n reputational damage and a public reversal — against the vendor's own commercial interest.\n- Harm predictable from their position? A deliberate plot predicts a cover-up, not a same-week\n public apology and rollback. The observed response is the OPPOSITE of what malice predicts.\n```\n\n### Step 4 — Choose starting posture · Step 5 — Set override signal · Step 6 — Hold prior\n\n```\n- Prior: emergent over-correction / honest engineering error (not malice)\n- Starting posture: treat as a capability + evaluation failure, not a conspiracy\n- First move: read the vendor's own post-incident statement; look for whether the fix is\n process/eval-level (consistent with error) or absent (would raise the malice prior)\n- Override trigger: internal documents or credible reporting showing the outputs were a\n KNOWN, intended result shipped deliberately — then update toward malice.\n```\n\n---\n\n## Case B: A competitor makes a surprising move\n\n**Anchor:** In early 2025, the Chinese lab DeepSeek released a high-performing open-weight model (DeepSeek-R1) that reportedly matched or approached frontier reasoning models at dramatically lower stated training cost. The release briefly rattled US markets and prompted some commentary framing it as a coordinated strategic strike, sabotage, or fabricated numbers designed to mislead rivals.\n\n### Step 1 — Describe the action and harm (factual)\n\n```\n- What was done: A competitor shipped a strong open-weight model and published a low cost figure.\n- Harm to me (an incumbent/investor): Competitive threat; a sharp, surprising market reaction.\n- Gut attribution: \"This is a deliberate psy-op / the numbers are a lie meant to hurt us.\"\n```\n\n### Step 2 — Construct the non-malice explanation\n\n```\n- Bad information they had: None required — a genuinely capable lab shipped a genuine result.\n- What you didn't realize: Efficient training and open-weight release are ordinary competitive\n strategies; a rival can simply be good at engineering.\n- Optimizing for: Their OWN adoption, mindshare, and talent draw — not primarily your harm.\n- Coverage: ~85%. \"A competent competitor made a strong, self-interested move\" explains the\n release fully; it does NOT require any plot against you.\n```\n\n### Step 3 — Name what malice would additionally require\n\n```\n- Info they'd need: A coordinated intent to manipulate specifically YOUR outcomes.\n- Motivation at your expense: Assumes your harm is the goal rather than a side effect of their gain.\n- Harm predictable from their position? A rival advancing its own position predicts exactly this\n move WITHOUT any conspiracy — the market reaction is a consequence, not necessarily an aim.\n```\n\n### Step 4–6 — Posture, override, hold\n\n```\n- Prior: ordinary competent competition (self-interest), not a targeted attack\n- Starting posture: benchmark the claim on its merits; compete, don't accuse\n- First move: independently verify the published figures rather than assuming fabrication;\n \"surprising\" is not \"dishonest\"\n- Override trigger: reproducible evidence the headline numbers were knowingly falsified, or\n documented coordination aimed at your specific harm — then update.\n```\n\n---\n\n## Why the razor is the right prior here — but only a prior\n\nModern AI systems make the mistake-prior *more* defensible than in ordinary human affairs, because they add failure modes that look intentional but are not:\n\n- **Emergent behavior.** Outputs arise from statistical training over vast data; harmful or biased results routinely appear that no engineer designed, chose, or foresaw. \"It happened\" does not imply \"someone wanted it to.\"\n- **The correction-layer interaction.** Guardrails, system prompts, and fine-tuning can interact with the base model to produce absurd results — an *accident of composition*, not a hidden message.\n- **Competitive surprise ≠ sabotage.** A rival's self-interested move can hurt you as a side effect while your harm is not the goal — the fundamental attribution error (Ross 1977) pushes observers to read intent into what is structural incentive.\n\nAnd the same guardrails from the SKILL apply. The razor is a **prior, not a proof**:\n\n- It does **not** excuse harm. A biased output is still harmful and still demands an evaluation/process fix — attribution is not accountability.\n- **Override on evidence.** If credible reporting or internal documents show a harmful output was a *known, intended* result shipped deliberately, or that a competitor knowingly falsified results to damage a specific rival, update toward malice.\n- **Watch the power/stakes asymmetry.** Where the cost of wrongly assuming non-malice is catastrophic (safety-critical deployment, deceptive-by-design products), shift toward a defensive posture *before* intent is proven.\n\n**Operational takeaway:** For 2024–2026 AI incidents, default to *emergent bug / honest error / ordinary self-interested competition* as the starting explanation, run the mistake-coverage estimate, name what malice would additionally require, and pre-commit an override signal. This produces a diagnostic conversation (root-cause, evals, verification) rather than a conspiracy narrative — while leaving the door open to update the moment real evidence of intent arrives.\n\n*Sources: Google, \"Gemini image generation got it wrong. We'll do better.\" (The Keyword / official Google blog, February 23, 2024); reporting on the Gemini image-generation pause, e.g., The Verge and BBC News (February 2024). DeepSeek, \"DeepSeek-R1\" release and technical report (January 2025); contemporaneous market-reaction reporting, e.g., Reuters and the Financial Times (late January 2025). Ross, L. (1977). \"The intuitive psychologist and his shortcomings,\" Advances in Experimental Social Psychology, 10, 173–220. Exact cost/benchmark figures are as publicly reported by the vendors and had not been independently verified at the time of the events.*\n\nFile v1.0.4:examples/hanlons-submission-1980-heinleins-1941-articulation.md\n\n# Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation\n\n> *Example for the [hanlons-razor](../SKILL.md) skill.*\n\nThe Hanlon's Razor formulation that spread across modern internet culture, business writing, and decision-theory literature was submitted by **Robert J. Hanlon** of Scranton, Pennsylvania, to the 1980 humor compendium *Murphy's Law Book Two: More Reasons Why Things Go Wrong*, edited by Arthur Bloch:\n\n> \"Never attribute to malice that which is adequately explained by stupidity.\"\n\n— Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two.* Price/Stern/Sloan. ISBN 978-0843105674.\n\nThe submission was one of dozens in a humorous compendium. Most have been forgotten. Hanlon's became one of the most-cited heuristics of the next 40 years — a fact that should itself give pause.\n\nThe deeper history runs back further. **Robert A. Heinlein**, in his 1941 novella \"Logic of Empire\" (published in *Astounding Science Fiction*, later collected in *The Past Through Tomorrow*, 1967), wrote a passage that historians of the saying believe Hanlon either independently re-discovered or consciously adapted:\n\n> \"You have attributed conditions to villainy that simply result from stupidity.\"\n\n— Heinlein, R. A. (1941). \"Logic of Empire.\" *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211, p. 224.\n\nThe \"Logic of Empire\" context is instructive. Heinlein's protagonists are abolitionists who have come to believe the slavery system on Venus is being maintained by an evil conspiracy. The older character corrects them: the system persists not because of conspiracy but because of structural incompetence — bad information, misaligned incentives, individual short-term rationality producing collective long-term harm. The point: **systems that look like conspiracies are usually structures that have evolved to produce the same outcome without anyone planning it**. This is Hanlon's razor applied at the level of institutions, not individuals.\n\nEven earlier, **Johann Wolfgang von Goethe** had written in *Sorrows of Young Werther* (1774):\n\n> \"Misunderstandings and lethargy perhaps produce more wrong in the world than deceit and malice do. At least the last two are certainly rarer.\"\n\n— Goethe, J. W. (1774). *Die Leiden des jungen Werthers.* As discussed at https://quoteinvestigator.com/2016/12/30/not-malice/\n\nThe three formulations span 200+ years, three languages, and quite different cultural contexts — yet articulate the same calibration error. The fact that the observation keeps being independently re-derived suggests a robust underlying psychological fact: **humans systematically over-attribute malice.**\n\nThe empirical psychology supports the heuristic. The **fundamental attribution error** (Ross 1977) documents that humans attribute others' behavior to disposition (intent, character) while attributing their own similar behavior to situation (circumstance, pressure). When someone cuts us off in traffic, \"they're a jerk\" (disposition); when we cut someone off, \"I had an emergency\" (situation). Hanlon's razor is a corrective: extend the situation-attribution we apply to ourselves to others' behavior.\n\n> \"The tendency for people to over-emphasize personal characteristics and ignore situational factors in judging others' behavior. ... Because of the fundamental attribution error, we tend to believe that others do bad things because they are bad people.\"\n>\n> — Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n\nThe framework has been adopted operationally:\n\n**Software engineering.** Post-mortems of production failures routinely apply Hanlon's razor: the engineer who pushed the bad commit was not malicious, and the corrective is process/tooling, not punishment. Google's **blameless postmortem** practice is Hanlon's razor institutionalized.\n\n**Corporate management.** Andy Grove, in *High Output Management*, applied the heuristic to performance management: a subordinate who keeps making the same mistake is most often not insubordinate — they are missing information, training, or context.\n\n**International diplomacy.** The \"rational actor\" framework in international relations is an application at the state level: states act in their own (sometimes incompetent) interest rather than out of pure malice.\n\nThe framework also has documented failure modes:\n\n**First, the razor can be weaponized.** People who have caused real harm can invoke Hanlon's razor to deflect accountability. The defense: the razor specifies a *starting prior*, not a final attribution. Repeated incompetence-causing-harm despite available information eventually becomes equivalent to malice in practical effect.\n\n**Second, the razor can disable necessary protective response.** When the cost of being wrong about non-malice is catastrophic (abusive partner, predatory company, hostile state), the calibrated prior must shift toward defensive posture *before* malice is proven.\n\n**Third, the razor doesn't excuse harm.** A bridge that collapses through incompetence has still collapsed. The point is *attribution*, not *accountability*. Once incompetence is identified, the corrective is competence-building, not blame-shifting.\n\nThree operational lessons from extensive application:\n\n**First, the conversation that follows from mistake-prior is almost always better than the one that follows from malice-prior — even when it later turns out to be malice.** Clarifying conversations produce information; accusatory ones produce defensiveness.\n\n**Second, the malice attribution feels viscerally correct in real-time and almost always feels overconfident in retrospect.** This calibration error is robust.\n\n**Third, the razor scales from individual to institutional.** Most \"bad culture\" / \"broken system\" complaints are about coordination failure, not coordinated malice.\n\nFile v1.0.4:skill-card.md\n\n## Description:
\nHelps an agent analyze situations where a user suspects malicious intent by first testing whether mistake, miscommunication, incompetence, or asymmetric information explains the evidence, while preserving clear override conditions.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[deciqai](https://clawhub.ai/user/deciqai)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nEmployees, teams, developers, and agents use this skill to de-escalate premature malice attribution in workplace, partner, company, or AI-incident contexts. It produces a structured Hanlon's Razor analysis with the factual harm, non-malice explanation, extra assumptions required for malice, starting posture, first action, and override signal.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: The skill can influence how an agent frames interpersonal or organizational conflict and may underweight malice in cases involving concrete evidence, abuse, severe power imbalance, or safety-critical consequences.
\nMitigation: Use it only as a starting prior, apply the skill's stated limits, and avoid it when the cost of wrongly assuming non-malice is catastrophic.
\nRisk: A generated analysis may sound more certain than the available facts support.
\nMitigation: Require a factual incident description, an explicit coverage estimate, and clear override signals before relying on the attribution.
\n\n\n## Reference(s):
\n- [Primary sources](references/sources.md)
\n- [Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)
\n- [AI Incidents: Incompetence and Emergent Error as a Prior Over Malice (2024-2026)](examples/ai-incidents-malice-vs-emergent-error-2024-2026.md)
\n- [Quote Investigator: Hanlon's Razor](https://quoteinvestigator.com/2016/12/30/not-malice/)
\n- [Google: Gemini image generation got it wrong](https://blog.google/products/gemini/gemini-image-generation-issue/)
\n\n\n## Skill Output:
\n**Output Type(s):** [text, markdown, guidance]
\n**Output Format:** [Markdown analysis with structured fields and short explanatory prose]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [Non-executable reasoning output; no shell commands, code changes, network access, or credential handling.]
\n\n## Skill Version(s):
\n1.0.4 (source: ClawHub release evidence)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v1.0.3: 5 files, 8718 bytes\n\nFiles: examples/hanlons-submission-1980-heinleins-1941-articulation.md (5953b), references/sources.md (881b), skill-card.md (2756b), SKILL.md (6380b), _meta.json (132b)\n\nFile v1.0.3:SKILL.md\n\n---\nname: hanlons-razor\ndescription: \"Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone is about to escalate based on assumed malicious intent; a pattern of bad outcomes is being labeled a coordinated attack. Do NOT activate when: concrete documented evidence of malice already exists; the cost of being wrong about non-malice is catastrophic (e.g., safety-critical or abusive-relationship context).\"\n---\n\n# Hanlon's Razor\n\n## Overview\n\nBefore assuming someone hurt you on purpose, construct the version where they made a mistake — and see how much evidence it explains. The razor is a Bayesian *prior*, not a proof; override it when concrete evidence of malice arrives. Human attribution systematically over-weights intent (fundamental attribution error); most hostile-seeming acts are incompetence, miscommunication, or asymmetric information.\n\nComposes with `bayesian-reasoning`, `abductive-reasoning`, `occams-razor`, `critical-thinking`.\n\n## When to Use\n\n- You feel an emotional pull toward \"they did this on purpose\"\n- You're about to escalate on the assumption of malice\n- A pattern of bad outcomes is being framed as a coordinated attack\n- A team is in conflict and each side believes the other is acting in bad faith\n\n**Not when:** concrete evidence of malicious intent exists; cost of being wrong is catastrophic; power imbalance makes \"they probably didn't mean it\" an abuse-enabling stance.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line: before believing someone did it on purpose, construct the mistake version — see how much evidence it covers.\n2. Check fit: concrete malice evidence / catastrophic cost of being wrong → not this lens.\n3. Elicit the specific incident — what exactly happened?\n> **[WAIT — do not advance until user responds]**\n4. Work through The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close: name the clarifying-conversation move + the override signal to watch for.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe the action and harm** (factual, not interpretive)\n```\n- What was done: \n- Harm to me: \n- Gut attribution: \n```\n**Step 2 — Construct the non-malice explanation**\n```\n- Bad information they had: / Didn't realize: / Optimizing for: / Under pressure from:\n- Coverage: <% of observed behavior this explains>\n```\n**Step 3 — Name what malice would additionally require**\n```\n- Info they'd need: / Motivation at your expense: / Harm predictable from their position?\n```\n**Step 4 — Choose starting posture** · **Step 5 — Set override signal** · **Step 6 — Hold prior until evidence changes it**\n```\n- Prior: · Starting posture: · First move: · Override trigger:\n```\n\n## Output Template\n```\n# Hanlon's Razor Analysis: \nMistake explanation + coverage % | Malice extra assumptions | Prior | Override signal | First action\n```\n\n*→ Method in Action: [Hanlon's Submission 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)*\n\n## Pack: Common Patterns\n\n| Situation | Default malice | Likely incompetence |\n|---|---|---|\n| Colleague sends harsh email | \"They're undermining me\" | Tired, didn't notice tone |\n| Manager ignores your work | \"They don't value me\" | Overloaded, missed the update |\n| Customer complains publicly | \"Trying to extort us\" | Real issue, tried other channels |\n| Investor passes after diligence | \"Were never serious\" | Portfolio rebalancing |\n| Co-founder makes unilateral decision | \"Cutting me out\" | Time pressure, assumed approval |\n\n## Applying It Well\n\n- Mistake-prior conversations produce information; malice-prior conversations produce defensiveness — even when malice later turns out true.\n- The malice attribution feels viscerally correct in real-time; almost always overconfident in retrospect.\n- Scales from individual to institutional: most \"broken culture\" is coordination failure, not coordinated malice.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] Apply razor then ignore mounting evidence of malice | It's a prior, not a permanent ban on updating. |\n| [D] Use razor to avoid hard conversations | It recommends a *clarifying* conversation, not silence. |\n| [D] Apply razor across power asymmetries | Boss-harming-subordinate: power shifts the calculus. |\n| [D] Invoke razor on yourself to dodge accountability | \"Didn't mean to\" doesn't undo the harm. |\n| [D] Stop investigating once incompetence is identified | Incompetence-caused harm still requires response. |\n| [D] Use razor to gaslight someone who was genuinely harmed | Attribution ≠ override of their lived experience. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Razor invoked by the actor, not the receiver\n- \"Incompetence\" explanation conveniently absolves a powerful party\n- Pattern of \"incompetence\" always harming the same direction\n- Razor used to silence a victim's report or applied repeatedly without updating priors\n\n## Verification\n\n- [ ] Action and harm stated factually (not interpretively)\n- [ ] Specific incompetence explanation constructed and coverage assessed\n- [ ] Additional assumptions required for malice named\n- [ ] Clarifying-conversation posture chosen (not accusation)\n- [ ] Override signals specified; power asymmetry considered\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 164 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/c/hanlons-razor** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\nFile v1.0.3:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"hanlons-razor\",\n \"version\": \"1.0.3\",\n \"publishedAt\": 1783508675507\n}\n\nFile v1.0.3:references/sources.md\n\n# Sources — hanlons-razor\n\n> *Primary sources for the [hanlons-razor](../SKILL.md) skill.*\n\n## Sources\n\n- Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two: More Reasons Why Things Go Wrong.* Price/Stern/Sloan. ISBN 978-0843105674.\n- Heinlein, R. A. (1941). \"Logic of Empire.\" In *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211.\n- Goethe, J. W. (1774). *Die Leiden des jungen Werthers.*\n- Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n- Akerlof, G. A. & Shiller, R. J. (2015). *Phishing for Phools.* Princeton University Press. ISBN 978-0691168319.\n- Grove, A. S. (1983). *High Output Management.* Random House. ISBN 978-0679762881.\n- \"Hanlon's Razor.\" *Quote Investigator.* https://quoteinvestigator.com/2016/12/30/not-malice/\n\nFile v1.0.3:examples/hanlons-submission-1980-heinleins-1941-articulation.md\n\n# Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation\n\n> *Example for the [hanlons-razor](../SKILL.md) skill.*\n\nThe Hanlon's Razor formulation that spread across modern internet culture, business writing, and decision-theory literature was submitted by **Robert J. Hanlon** of Scranton, Pennsylvania, to the 1980 humor compendium *Murphy's Law Book Two: More Reasons Why Things Go Wrong*, edited by Arthur Bloch:\n\n> \"Never attribute to malice that which is adequately explained by stupidity.\"\n\n— Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two.* Price/Stern/Sloan. ISBN 978-0843105674.\n\nThe submission was one of dozens in a humorous compendium. Most have been forgotten. Hanlon's became one of the most-cited heuristics of the next 40 years — a fact that should itself give pause.\n\nThe deeper history runs back further. **Robert A. Heinlein**, in his 1941 novella \"Logic of Empire\" (published in *Astounding Science Fiction*, later collected in *The Past Through Tomorrow*, 1967), wrote a passage that historians of the saying believe Hanlon either independently re-discovered or consciously adapted:\n\n> \"You have attributed conditions to villainy that simply result from stupidity.\"\n\n— Heinlein, R. A. (1941). \"Logic of Empire.\" *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211, p. 224.\n\nThe \"Logic of Empire\" context is instructive. Heinlein's protagonists are abolitionists who have come to believe the slavery system on Venus is being maintained by an evil conspiracy. The older character corrects them: the system persists not because of conspiracy but because of structural incompetence — bad information, misaligned incentives, individual short-term rationality producing collective long-term harm. The point: **systems that look like conspiracies are usually structures that have evolved to produce the same outcome without anyone planning it**. This is Hanlon's razor applied at the level of institutions, not individuals.\n\nEven earlier, **Johann Wolfgang von Goethe** had written in *Sorrows of Young Werther* (1774):\n\n> \"Misunderstandings and lethargy perhaps produce more wrong in the world than deceit and malice do. At least the last two are certainly rarer.\"\n\n— Goethe, J. W. (1774). *Die Leiden des jungen Werthers.* As discussed at https://quoteinvestigator.com/2016/12/30/not-malice/\n\nThe three formulations span 200+ years, three languages, and quite different cultural contexts — yet articulate the same calibration error. The fact that the observation keeps being independently re-derived suggests a robust underlying psychological fact: **humans systematically over-attribute malice.**\n\nThe empirical psychology supports the heuristic. The **fundamental attribution error** (Ross 1977) documents that humans attribute others' behavior to disposition (intent, character) while attributing their own similar behavior to situation (circumstance, pressure). When someone cuts us off in traffic, \"they're a jerk\" (disposition); when we cut someone off, \"I had an emergency\" (situation). Hanlon's razor is a corrective: extend the situation-attribution we apply to ourselves to others' behavior.\n\n> \"The tendency for people to over-emphasize personal characteristics and ignore situational factors in judging others' behavior. ... Because of the fundamental attribution error, we tend to believe that others do bad things because they are bad people.\"\n>\n> — Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n\nThe framework has been adopted operationally:\n\n**Software engineering.** Post-mortems of production failures routinely apply Hanlon's razor: the engineer who pushed the bad commit was not malicious, and the corrective is process/tooling, not punishment. Google's **blameless postmortem** practice is Hanlon's razor institutionalized.\n\n**Corporate management.** Andy Grove, in *High Output Management*, applied the heuristic to performance management: a subordinate who keeps making the same mistake is most often not insubordinate — they are missing information, training, or context.\n\n**International diplomacy.** The \"rational actor\" framework in international relations is an application at the state level: states act in their own (sometimes incompetent) interest rather than out of pure malice.\n\nThe framework also has documented failure modes:\n\n**First, the razor can be weaponized.** People who have caused real harm can invoke Hanlon's razor to deflect accountability. The defense: the razor specifies a *starting prior*, not a final attribution. Repeated incompetence-causing-harm despite available information eventually becomes equivalent to malice in practical effect.\n\n**Second, the razor can disable necessary protective response.** When the cost of being wrong about non-malice is catastrophic (abusive partner, predatory company, hostile state), the calibrated prior must shift toward defensive posture *before* malice is proven.\n\n**Third, the razor doesn't excuse harm.** A bridge that collapses through incompetence has still collapsed. The point is *attribution*, not *accountability*. Once incompetence is identified, the corrective is competence-building, not blame-shifting.\n\nThree operational lessons from extensive application:\n\n**First, the conversation that follows from mistake-prior is almost always better than the one that follows from malice-prior — even when it later turns out to be malice.** Clarifying conversations produce information; accusatory ones produce defensiveness.\n\n**Second, the malice attribution feels viscerally correct in real-time and almost always feels overconfident in retrospect.** This calibration error is robust.\n\n**Third, the razor scales from individual to institutional.** Most \"bad culture\" / \"broken system\" complaints are about coordination failure, not coordinated malice.\n\nFile v1.0.3:skill-card.md\n\n## Description:
\nApplies Hanlon's Razor to help an agent test non-malicious explanations before escalating on assumed bad intent, while preserving explicit override conditions for documented malice, catastrophic downside, and power imbalance.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[deciqai](https://clawhub.ai/user/deciqai)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nEmployees, operators, and agents use this skill to analyze interpersonal, team, customer, or institutional conflict before assuming malicious intent. It guides a factual incident description, a plausible mistake-based explanation, explicit malice assumptions, an override signal, and a first clarifying action.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: The skill can bias an agent toward non-malicious explanations in abuse, safety-critical, or other high-stakes contexts.
\nMitigation: Do not use this lens when concrete malice evidence exists or the downside of assuming non-malice is catastrophic; shift to a protective posture instead.
\nRisk: A powerful actor or repeated harmful pattern could be excused as mere incompetence.
\nMitigation: Require explicit override signals, consider power imbalance, and keep investigating when the same pattern repeatedly harms the same party.
\nRisk: A mistake-prior framing could discourage accountability after real harm.
\nMitigation: Separate attribution from accountability: use the skill to choose the first conversation posture, not to dismiss the harm or avoid corrective action.
\n\n\n## Reference(s):
\n- [ClawHub skill page](https://clawhub.ai/deciqai/skills/hanlons-razor)
\n- [Sources - hanlons-razor](references/sources.md)
\n- [Method in Action: Hanlon's Submission 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)
\n- [Quote Investigator: Hanlon's Razor](https://quoteinvestigator.com/2016/12/30/not-malice/)
\n\n\n## Skill Output:
\n**Output Type(s):** [text, markdown, guidance, analysis]
\n**Output Format:** [Markdown analysis with structured prompts and a concise incident assessment]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [Text-only reasoning output; no scripts, credentials, persistence, or external tools.]
\n\n## Skill Version(s):
\n1.0.3 (source: server release evidence)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v1.0.2: 5 files, 8641 bytes\n\nFiles: examples/hanlons-submission-1980-heinleins-1941-articulation.md (5953b), references/sources.md (881b), skill-card.md (2411b), SKILL.md (6483b), _meta.json (132b)\n\nFile v1.0.2:SKILL.md\n\n---\nname: hanlons-razor\ndescription: \"Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone is about to escalate based on assumed malicious intent; a pattern of bad outcomes is being labeled a coordinated attack. Do NOT activate when: concrete documented evidence of malice already exists; the cost of being wrong about non-malice is catastrophic (e.g., safety-critical or abusive-relationship context).\"\n---\n\n# Hanlon's Razor\n\n## Overview\n\nBefore assuming someone hurt you on purpose, construct the version where they made a mistake — and see how much evidence it explains. The razor is a Bayesian *prior*, not a proof; override it when concrete evidence of malice arrives. Human attribution systematically over-weights intent (fundamental attribution error); most hostile-seeming acts are incompetence, miscommunication, or asymmetric information.\n\nComposes with `bayesian-reasoning`, `abductive-reasoning`, `occams-razor`, `critical-thinking`.\n\n## When to Use\n\n- You feel an emotional pull toward \"they did this on purpose\"\n- You're about to escalate on the assumption of malice\n- A pattern of bad outcomes is being framed as a coordinated attack\n- A team is in conflict and each side believes the other is acting in bad faith\n\n**Not when:** concrete evidence of malicious intent exists; cost of being wrong is catastrophic; power imbalance makes \"they probably didn't mean it\" an abuse-enabling stance.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line: before believing someone did it on purpose, construct the mistake version — see how much evidence it covers.\n2. Check fit: concrete malice evidence / catastrophic cost of being wrong → not this lens.\n3. Elicit the specific incident — what exactly happened?\n> **[WAIT — do not advance until user responds]**\n4. Work through The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close: name the clarifying-conversation move + the override signal to watch for.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe the action and harm** (factual, not interpretive)\n```\n- What was done: \n- Harm to me: \n- Gut attribution: \n```\n**Step 2 — Construct the non-malice explanation**\n```\n- Bad information they had: / Didn't realize: / Optimizing for: / Under pressure from:\n- Coverage: <% of observed behavior this explains>\n```\n**Step 3 — Name what malice would additionally require**\n```\n- Info they'd need: / Motivation at your expense: / Harm predictable from their position?\n```\n**Step 4 — Choose starting posture** · **Step 5 — Set override signal** · **Step 6 — Hold prior until evidence changes it**\n```\n- Prior: · Starting posture: · First move: · Override trigger:\n```\n\n## Output Template\n```\n# Hanlon's Razor Analysis: \nMistake explanation + coverage % | Malice extra assumptions | Prior | Override signal | First action\n```\n\n*→ Method in Action: [Hanlon's Submission 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)*\n\n## Pack: Common Patterns\n\n| Situation | Default malice | Likely incompetence |\n|---|---|---|\n| Colleague sends harsh email | \"They're undermining me\" | Tired, didn't notice tone |\n| Manager ignores your work | \"They don't value me\" | Overloaded, missed the update |\n| Customer complains publicly | \"Trying to extort us\" | Real issue, tried other channels |\n| Investor passes after diligence | \"Were never serious\" | Portfolio rebalancing |\n| Co-founder makes unilateral decision | \"Cutting me out\" | Time pressure, assumed approval |\n\n## Applying It Well\n\n- Mistake-prior conversations produce information; malice-prior conversations produce defensiveness — even when malice later turns out true.\n- The malice attribution feels viscerally correct in real-time; almost always overconfident in retrospect.\n- Scales from individual to institutional: most \"broken culture\" is coordination failure, not coordinated malice.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] Apply razor then ignore mounting evidence of malice | It's a prior, not a permanent ban on updating. |\n| [D] Use razor to avoid hard conversations | It recommends a *clarifying* conversation, not silence. |\n| [D] Apply razor across power asymmetries | Boss-harming-subordinate: power shifts the calculus. |\n| [D] Invoke razor on yourself to dodge accountability | \"Didn't mean to\" doesn't undo the harm. |\n| [D] Stop investigating once incompetence is identified | Incompetence-caused harm still requires response. |\n| [D] Use razor to gaslight someone who was genuinely harmed | Attribution ≠ override of their lived experience. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Razor invoked by the actor, not the receiver\n- \"Incompetence\" explanation conveniently absolves a powerful party\n- Pattern of \"incompetence\" always harming the same direction\n- Razor used to silence a victim's report or applied repeatedly without updating priors\n\n## Verification\n\n- [ ] Action and harm stated factually (not interpretively)\n- [ ] Specific incompetence explanation constructed and coverage assessed\n- [ ] Additional assumptions required for malice named\n- [ ] Clarifying-conversation posture chosen (not accusation)\n- [ ] Override signals specified; power asymmetry considered\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 163 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/skills/hanlons-razor?utm_source=clawhub&utm_medium=marketplace&utm_campaign=knowledge-skills&utm_content=hanlons-razor** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\nFile v1.0.2:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"hanlons-razor\",\n \"version\": \"1.0.2\",\n \"publishedAt\": 1783471780182\n}\n\nFile v1.0.2:references/sources.md\n\n# Sources — hanlons-razor\n\n> *Primary sources for the [hanlons-razor](../SKILL.md) skill.*\n\n## Sources\n\n- Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two: More Reasons Why Things Go Wrong.* Price/Stern/Sloan. ISBN 978-0843105674.\n- Heinlein, R. A. (1941). \"Logic of Empire.\" In *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211.\n- Goethe, J. W. (1774). *Die Leiden des jungen Werthers.*\n- Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n- Akerlof, G. A. & Shiller, R. J. (2015). *Phishing for Phools.* Princeton University Press. ISBN 978-0691168319.\n- Grove, A. S. (1983). *High Output Management.* Random House. ISBN 978-0679762881.\n- \"Hanlon's Razor.\" *Quote Investigator.* https://quoteinvestigator.com/2016/12/30/not-malice/\n\nFile v1.0.2:examples/hanlons-submission-1980-heinleins-1941-articulation.md\n\n# Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation\n\n> *Example for the [hanlons-razor](../SKILL.md) skill.*\n\nThe Hanlon's Razor formulation that spread across modern internet culture, business writing, and decision-theory literature was submitted by **Robert J. Hanlon** of Scranton, Pennsylvania, to the 1980 humor compendium *Murphy's Law Book Two: More Reasons Why Things Go Wrong*, edited by Arthur Bloch:\n\n> \"Never attribute to malice that which is adequately explained by stupidity.\"\n\n— Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two.* Price/Stern/Sloan. ISBN 978-0843105674.\n\nThe submission was one of dozens in a humorous compendium. Most have been forgotten. Hanlon's became one of the most-cited heuristics of the next 40 years — a fact that should itself give pause.\n\nThe deeper history runs back further. **Robert A. Heinlein**, in his 1941 novella \"Logic of Empire\" (published in *Astounding Science Fiction*, later collected in *The Past Through Tomorrow*, 1967), wrote a passage that historians of the saying believe Hanlon either independently re-discovered or consciously adapted:\n\n> \"You have attributed conditions to villainy that simply result from stupidity.\"\n\n— Heinlein, R. A. (1941). \"Logic of Empire.\" *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211, p. 224.\n\nThe \"Logic of Empire\" context is instructive. Heinlein's protagonists are abolitionists who have come to believe the slavery system on Venus is being maintained by an evil conspiracy. The older character corrects them: the system persists not because of conspiracy but because of structural incompetence — bad information, misaligned incentives, individual short-term rationality producing collective long-term harm. The point: **systems that look like conspiracies are usually structures that have evolved to produce the same outcome without anyone planning it**. This is Hanlon's razor applied at the level of institutions, not individuals.\n\nEven earlier, **Johann Wolfgang von Goethe** had written in *Sorrows of Young Werther* (1774):\n\n> \"Misunderstandings and lethargy perhaps produce more wrong in the world than deceit and malice do. At least the last two are certainly rarer.\"\n\n— Goethe, J. W. (1774). *Die Leiden des jungen Werthers.* As discussed at https://quoteinvestigator.com/2016/12/30/not-malice/\n\nThe three formulations span 200+ years, three languages, and quite different cultural contexts — yet articulate the same calibration error. The fact that the observation keeps being independently re-derived suggests a robust underlying psychological fact: **humans systematically over-attribute malice.**\n\nThe empirical psychology supports the heuristic. The **fundamental attribution error** (Ross 1977) documents that humans attribute others' behavior to disposition (intent, character) while attributing their own similar behavior to situation (circumstance, pressure). When someone cuts us off in traffic, \"they're a jerk\" (disposition); when we cut someone off, \"I had an emergency\" (situation). Hanlon's razor is a corrective: extend the situation-attribution we apply to ourselves to others' behavior.\n\n> \"The tendency for people to over-emphasize personal characteristics and ignore situational factors in judging others' behavior. ... Because of the fundamental attribution error, we tend to believe that others do bad things because they are bad people.\"\n>\n> — Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n\nThe framework has been adopted operationally:\n\n**Software engineering.** Post-mortems of production failures routinely apply Hanlon's razor: the engineer who pushed the bad commit was not malicious, and the corrective is process/tooling, not punishment. Google's **blameless postmortem** practice is Hanlon's razor institutionalized.\n\n**Corporate management.** Andy Grove, in *High Output Management*, applied the heuristic to performance management: a subordinate who keeps making the same mistake is most often not insubordinate — they are missing information, training, or context.\n\n**International diplomacy.** The \"rational actor\" framework in international relations is an application at the state level: states act in their own (sometimes incompetent) interest rather than out of pure malice.\n\nThe framework also has documented failure modes:\n\n**First, the razor can be weaponized.** People who have caused real harm can invoke Hanlon's razor to deflect accountability. The defense: the razor specifies a *starting prior*, not a final attribution. Repeated incompetence-causing-harm despite available information eventually becomes equivalent to malice in practical effect.\n\n**Second, the razor can disable necessary protective response.** When the cost of being wrong about non-malice is catastrophic (abusive partner, predatory company, hostile state), the calibrated prior must shift toward defensive posture *before* malice is proven.\n\n**Third, the razor doesn't excuse harm.** A bridge that collapses through incompetence has still collapsed. The point is *attribution*, not *accountability*. Once incompetence is identified, the corrective is competence-building, not blame-shifting.\n\nThree operational lessons from extensive application:\n\n**First, the conversation that follows from mistake-prior is almost always better than the one that follows from malice-prior — even when it later turns out to be malice.** Clarifying conversations produce information; accusatory ones produce defensiveness.\n\n**Second, the malice attribution feels viscerally correct in real-time and almost always feels overconfident in retrospect.** This calibration error is robust.\n\n**Third, the razor scales from individual to institutional.** Most \"bad culture\" / \"broken system\" complaints are about coordination failure, not coordinated malice.\n\nFile v1.0.2:skill-card.md\n\n## Description:
\nHelps an agent assess suspected bad-faith actions by first constructing a non-malice explanation, identifying what malice would additionally require, and setting override signals.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[deciqai](https://clawhub.ai/user/deciqai)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nEmployees, teams, and facilitators use this skill to examine emotionally charged conflict or escalation scenarios before concluding that another person or group acted maliciously. It guides a clarifying first posture while preserving explicit override conditions for concrete malice evidence, catastrophic downside, and power asymmetry.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: The skill could be misapplied where documented malice exists, where the downside of assuming non-malice is catastrophic, or where power imbalance makes non-malice framing unsafe.
\nMitigation: Apply the skill's fit check first; choose a protective posture instead of this lens when those override conditions are present.
\nRisk: A mistake-prior analysis could be used to excuse harm or stop investigation too early.
\nMitigation: Treat the razor as a starting prior, preserve accountability for harm, and define concrete override signals before acting.
\n\n\n## Reference(s):
\n- [Sources - hanlons-razor](references/sources.md)
\n- [Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)
\n- [Hanlon's Razor - Quote Investigator](https://quoteinvestigator.com/2016/12/30/not-malice/)
\n\n\n## Skill Output:
\n**Output Type(s):** [text, markdown, guidance]
\n**Output Format:** [Markdown analysis with structured prompts and concise recommendations]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [May pause for user input during coach-mode steps before completing the analysis.]
\n\n## Skill Version(s):
\n1.0.2 (source: server release evidence)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v1.0.1: 5 files, 8512 bytes\n\nFiles: examples/hanlons-submission-1980-heinleins-1941-articulation.md (5953b), references/sources.md (881b), skill-card.md (2389b), SKILL.md (6359b), _meta.json (132b)\n\nFile v1.0.1:SKILL.md\n\n---\nname: hanlons-razor\ndescription: \"Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone is about to escalate based on assumed malicious intent; a pattern of bad outcomes is being labeled a coordinated attack. Do NOT activate when: concrete documented evidence of malice already exists; the cost of being wrong about non-malice is catastrophic (e.g., safety-critical or abusive-relationship context).\"\n---\n\n# Hanlon's Razor\n\n## Overview\n\nBefore assuming someone hurt you on purpose, construct the version where they made a mistake — and see how much evidence it explains. The razor is a Bayesian *prior*, not a proof; override it when concrete evidence of malice arrives. Human attribution systematically over-weights intent (fundamental attribution error); most hostile-seeming acts are incompetence, miscommunication, or asymmetric information.\n\nComposes with [`bayesian-reasoning`](../bayesian-reasoning/SKILL.md), [`abductive-reasoning`](../abductive-reasoning/SKILL.md), [`occams-razor`](../occams-razor/SKILL.md), [`critical-thinking`](../critical-thinking/SKILL.md).\n\n## When to Use\n\n- You feel an emotional pull toward \"they did this on purpose\"\n- You're about to escalate on the assumption of malice\n- A pattern of bad outcomes is being framed as a coordinated attack\n- A team is in conflict and each side believes the other is acting in bad faith\n\n**Not when:** concrete evidence of malicious intent exists; cost of being wrong is catastrophic; power imbalance makes \"they probably didn't mean it\" an abuse-enabling stance.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line: before believing someone did it on purpose, construct the mistake version — see how much evidence it covers.\n2. Check fit: concrete malice evidence / catastrophic cost of being wrong → not this lens.\n3. Elicit the specific incident — what exactly happened?\n> **[WAIT — do not advance until user responds]**\n4. Work through The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close: name the clarifying-conversation move + the override signal to watch for.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe the action and harm** (factual, not interpretive)\n```\n- What was done: \n- Harm to me: \n- Gut attribution: \n```\n**Step 2 — Construct the non-malice explanation**\n```\n- Bad information they had: / Didn't realize: / Optimizing for: / Under pressure from:\n- Coverage: <% of observed behavior this explains>\n```\n**Step 3 — Name what malice would additionally require**\n```\n- Info they'd need: / Motivation at your expense: / Harm predictable from their position?\n```\n**Step 4 — Choose starting posture** · **Step 5 — Set override signal** · **Step 6 — Hold prior until evidence changes it**\n```\n- Prior: · Starting posture: · First move: · Override trigger:\n```\n\n## Output Template\n```\n# Hanlon's Razor Analysis: \nMistake explanation + coverage % | Malice extra assumptions | Prior | Override signal | First action\n```\n\n*→ Method in Action: [Hanlon's Submission 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)*\n\n## Pack: Common Patterns\n\n| Situation | Default malice | Likely incompetence |\n|---|---|---|\n| Colleague sends harsh email | \"They're undermining me\" | Tired, didn't notice tone |\n| Manager ignores your work | \"They don't value me\" | Overloaded, missed the update |\n| Customer complains publicly | \"Trying to extort us\" | Real issue, tried other channels |\n| Investor passes after diligence | \"Were never serious\" | Portfolio rebalancing |\n| Co-founder makes unilateral decision | \"Cutting me out\" | Time pressure, assumed approval |\n\n## Applying It Well\n\n- Mistake-prior conversations produce information; malice-prior conversations produce defensiveness — even when malice later turns out true.\n- The malice attribution feels viscerally correct in real-time; almost always overconfident in retrospect.\n- Scales from individual to institutional: most \"broken culture\" is coordination failure, not coordinated malice.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] Apply razor then ignore mounting evidence of malice | It's a prior, not a permanent ban on updating. |\n| [D] Use razor to avoid hard conversations | It recommends a *clarifying* conversation, not silence. |\n| [D] Apply razor across power asymmetries | Boss-harming-subordinate: power shifts the calculus. |\n| [D] Invoke razor on yourself to dodge accountability | \"Didn't mean to\" doesn't undo the harm. |\n| [D] Stop investigating once incompetence is identified | Incompetence-caused harm still requires response. |\n| [D] Use razor to gaslight someone who was genuinely harmed | Attribution ≠ override of their lived experience. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Razor invoked by the actor, not the receiver\n- \"Incompetence\" explanation conveniently absolves a powerful party\n- Pattern of \"incompetence\" always harming the same direction\n- Razor used to silence a victim's report or applied repeatedly without updating priors\n\n## Verification\n\n- [ ] Action and harm stated factually (not interpretively)\n- [ ] Specific incompetence explanation constructed and coverage assessed\n- [ ] Additional assumptions required for malice named\n- [ ] Clarifying-conversation posture chosen (not accusation)\n- [ ] Override signals specified; power asymmetry considered\n\n---\n\n*Part of **deciqAI Knowledge Skills** — open-source thinking skills that make rigor executable for AI agents. Built by deciqAI · https://deciqai.com · Contributions welcome — see the template at the repo root.*\n\nFile v1.0.1:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"hanlons-razor\",\n \"version\": \"1.0.1\",\n \"publishedAt\": 1783456427523\n}\n\nFile v1.0.1:references/sources.md\n\n# Sources — hanlons-razor\n\n> *Primary sources for the [hanlons-razor](../SKILL.md) skill.*\n\n## Sources\n\n- Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two: More Reasons Why Things Go Wrong.* Price/Stern/Sloan. ISBN 978-0843105674.\n- Heinlein, R. A. (1941). \"Logic of Empire.\" In *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211.\n- Goethe, J. W. (1774). *Die Leiden des jungen Werthers.*\n- Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n- Akerlof, G. A. & Shiller, R. J. (2015). *Phishing for Phools.* Princeton University Press. ISBN 978-0691168319.\n- Grove, A. S. (1983). *High Output Management.* Random House. ISBN 978-0679762881.\n- \"Hanlon's Razor.\" *Quote Investigator.* https://quoteinvestigator.com/2016/12/30/not-malice/\n\nFile v1.0.1:examples/hanlons-submission-1980-heinleins-1941-articulation.md\n\n# Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation\n\n> *Example for the [hanlons-razor](../SKILL.md) skill.*\n\nThe Hanlon's Razor formulation that spread across modern internet culture, business writing, and decision-theory literature was submitted by **Robert J. Hanlon** of Scranton, Pennsylvania, to the 1980 humor compendium *Murphy's Law Book Two: More Reasons Why Things Go Wrong*, edited by Arthur Bloch:\n\n> \"Never attribute to malice that which is adequately explained by stupidity.\"\n\n— Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two.* Price/Stern/Sloan. ISBN 978-0843105674.\n\nThe submission was one of dozens in a humorous compendium. Most have been forgotten. Hanlon's became one of the most-cited heuristics of the next 40 years — a fact that should itself give pause.\n\nThe deeper history runs back further. **Robert A. Heinlein**, in his 1941 novella \"Logic of Empire\" (published in *Astounding Science Fiction*, later collected in *The Past Through Tomorrow*, 1967), wrote a passage that historians of the saying believe Hanlon either independently re-discovered or consciously adapted:\n\n> \"You have attributed conditions to villainy that simply result from stupidity.\"\n\n— Heinlein, R. A. (1941). \"Logic of Empire.\" *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211, p. 224.\n\nThe \"Logic of Empire\" context is instructive. Heinlein's protagonists are abolitionists who have come to believe the slavery system on Venus is being maintained by an evil conspiracy. The older character corrects them: the system persists not because of conspiracy but because of structural incompetence — bad information, misaligned incentives, individual short-term rationality producing collective long-term harm. The point: **systems that look like conspiracies are usually structures that have evolved to produce the same outcome without anyone planning it**. This is Hanlon's razor applied at the level of institutions, not individuals.\n\nEven earlier, **Johann Wolfgang von Goethe** had written in *Sorrows of Young Werther* (1774):\n\n> \"Misunderstandings and lethargy perhaps produce more wrong in the world than deceit and malice do. At least the last two are certainly rarer.\"\n\n— Goethe, J. W. (1774). *Die Leiden des jungen Werthers.* As discussed at https://quoteinvestigator.com/2016/12/30/not-malice/\n\nThe three formulations span 200+ years, three languages, and quite different cultural contexts — yet articulate the same calibration error. The fact that the observation keeps being independently re-derived suggests a robust underlying psychological fact: **humans systematically over-attribute malice.**\n\nThe empirical psychology supports the heuristic. The **fundamental attribution error** (Ross 1977) documents that humans attribute others' behavior to disposition (intent, character) while attributing their own similar behavior to situation (circumstance, pressure). When someone cuts us off in traffic, \"they're a jerk\" (disposition); when we cut someone off, \"I had an emergency\" (situation). Hanlon's razor is a corrective: extend the situation-attribution we apply to ourselves to others' behavior.\n\n> \"The tendency for people to over-emphasize personal characteristics and ignore situational factors in judging others' behavior. ... Because of the fundamental attribution error, we tend to believe that others do bad things because they are bad people.\"\n>\n> — Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n\nThe framework has been adopted operationally:\n\n**Software engineering.** Post-mortems of production failures routinely apply Hanlon's razor: the engineer who pushed the bad commit was not malicious, and the corrective is process/tooling, not punishment. Google's **blameless postmortem** practice is Hanlon's razor institutionalized.\n\n**Corporate management.** Andy Grove, in *High Output Management*, applied the heuristic to performance management: a subordinate who keeps making the same mistake is most often not insubordinate — they are missing information, training, or context.\n\n**International diplomacy.** The \"rational actor\" framework in international relations is an application at the state level: states act in their own (sometimes incompetent) interest rather than out of pure malice.\n\nThe framework also has documented failure modes:\n\n**First, the razor can be weaponized.** People who have caused real harm can invoke Hanlon's razor to deflect accountability. The defense: the razor specifies a *starting prior*, not a final attribution. Repeated incompetence-causing-harm despite available information eventually becomes equivalent to malice in practical effect.\n\n**Second, the razor can disable necessary protective response.** When the cost of being wrong about non-malice is catastrophic (abusive partner, predatory company, hostile state), the calibrated prior must shift toward defensive posture *before* malice is proven.\n\n**Third, the razor doesn't excuse harm.** A bridge that collapses through incompetence has still collapsed. The point is *attribution*, not *accountability*. Once incompetence is identified, the corrective is competence-building, not blame-shifting.\n\nThree operational lessons from extensive application:\n\n**First, the conversation that follows from mistake-prior is almost always better than the one that follows from malice-prior — even when it later turns out to be malice.** Clarifying conversations produce information; accusatory ones produce defensiveness.\n\n**Second, the malice attribution feels viscerally correct in real-time and almost always feels overconfident in retrospect.** This calibration error is robust.\n\n**Third, the razor scales from individual to institutional.** Most \"bad culture\" / \"broken system\" complaints are about coordination failure, not coordinated malice.\n\nFile v1.0.1:skill-card.md\n\n## Description:
\nGuides agents through Hanlon's Razor to test non-malicious explanations before escalating assumed bad intent, while preserving override signals for concrete malice, serious safety risk, or abuse-enabling power imbalance.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[deciqai](https://clawhub.ai/user/deciqai)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nExternal users, employees, and developers use this skill to analyze conflicts or harmful incidents by comparing non-malicious explanations against the evidence required for malice. It supports calmer first actions while identifying override signals for documented malice, catastrophic downside, or power imbalance.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: Users may treat the skill as advice to dismiss real harm or delay protective action.
\nMitigation: Use it only as a starting prior; switch posture when there is concrete evidence of malice, abuse, catastrophic downside, or a power imbalance that makes non-malice assumptions unsafe.
\nRisk: The skill can understate accountability when harm is caused by incompetence rather than intent.
\nMitigation: Keep attribution separate from response: incompetence-caused harm still requires investigation, correction, and appropriate accountability.
\n\n\n## Reference(s):
\n- [Primary Sources](references/sources.md)
\n- [Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)
\n- [Quote Investigator: Hanlon's Razor](https://quoteinvestigator.com/2016/12/30/not-malice/)
\n\n\n## Skill Output:
\n**Output Type(s):** [text, markdown, guidance]
\n**Output Format:** [Markdown analysis template with concise reasoning steps]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [May ask step-by-step clarification questions before producing the final analysis.]
\n\n## Skill Version(s):
\n1.0.1 (source: release evidence)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v1.0.0: 5 files, 8410 bytes\n\nFiles: examples/hanlons-submission-1980-heinleins-1941-articulation.md (5953b), references/sources.md (881b), skill-card.md (2107b), SKILL.md (6359b), _meta.json (132b)\n\nFile v1.0.0:SKILL.md\n\n---\nname: hanlons-razor\ndescription: \"Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone is about to escalate based on assumed malicious intent; a pattern of bad outcomes is being labeled a coordinated attack. Do NOT activate when: concrete documented evidence of malice already exists; the cost of being wrong about non-malice is catastrophic (e.g., safety-critical or abusive-relationship context).\"\n---\n\n# Hanlon's Razor\n\n## Overview\n\nBefore assuming someone hurt you on purpose, construct the version where they made a mistake — and see how much evidence it explains. The razor is a Bayesian *prior*, not a proof; override it when concrete evidence of malice arrives. Human attribution systematically over-weights intent (fundamental attribution error); most hostile-seeming acts are incompetence, miscommunication, or asymmetric information.\n\nComposes with [`bayesian-reasoning`](../bayesian-reasoning/SKILL.md), [`abductive-reasoning`](../abductive-reasoning/SKILL.md), [`occams-razor`](../occams-razor/SKILL.md), [`critical-thinking`](../critical-thinking/SKILL.md).\n\n## When to Use\n\n- You feel an emotional pull toward \"they did this on purpose\"\n- You're about to escalate on the assumption of malice\n- A pattern of bad outcomes is being framed as a coordinated attack\n- A team is in conflict and each side believes the other is acting in bad faith\n\n**Not when:** concrete evidence of malicious intent exists; cost of being wrong is catastrophic; power imbalance makes \"they probably didn't mean it\" an abuse-enabling stance.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line: before believing someone did it on purpose, construct the mistake version — see how much evidence it covers.\n2. Check fit: concrete malice evidence / catastrophic cost of being wrong → not this lens.\n3. Elicit the specific incident — what exactly happened?\n> **[WAIT — do not advance until user responds]**\n4. Work through The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close: name the clarifying-conversation move + the override signal to watch for.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe the action and harm** (factual, not interpretive)\n```\n- What was done: \n- Harm to me: \n- Gut attribution: \n```\n**Step 2 — Construct the non-malice explanation**\n```\n- Bad information they had: / Didn't realize: / Optimizing for: / Under pressure from:\n- Coverage: <% of observed behavior this explains>\n```\n**Step 3 — Name what malice would additionally require**\n```\n- Info they'd need: / Motivation at your expense: / Harm predictable from their position?\n```\n**Step 4 — Choose starting posture** · **Step 5 — Set override signal** · **Step 6 — Hold prior until evidence changes it**\n```\n- Prior: · Starting posture: · First move: · Override trigger:\n```\n\n## Output Template\n```\n# Hanlon's Razor Analysis: \nMistake explanation + coverage % | Malice extra assumptions | Prior | Override signal | First action\n```\n\n*→ Method in Action: [Hanlon's Submission 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)*\n\n## Pack: Common Patterns\n\n| Situation | Default malice | Likely incompetence |\n|---|---|---|\n| Colleague sends harsh email | \"They're undermining me\" | Tired, didn't notice tone |\n| Manager ignores your work | \"They don't value me\" | Overloaded, missed the update |\n| Customer complains publicly | \"Trying to extort us\" | Real issue, tried other channels |\n| Investor passes after diligence | \"Were never serious\" | Portfolio rebalancing |\n| Co-founder makes unilateral decision | \"Cutting me out\" | Time pressure, assumed approval |\n\n## Applying It Well\n\n- Mistake-prior conversations produce information; malice-prior conversations produce defensiveness — even when malice later turns out true.\n- The malice attribution feels viscerally correct in real-time; almost always overconfident in retrospect.\n- Scales from individual to institutional: most \"broken culture\" is coordination failure, not coordinated malice.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] Apply razor then ignore mounting evidence of malice | It's a prior, not a permanent ban on updating. |\n| [D] Use razor to avoid hard conversations | It recommends a *clarifying* conversation, not silence. |\n| [D] Apply razor across power asymmetries | Boss-harming-subordinate: power shifts the calculus. |\n| [D] Invoke razor on yourself to dodge accountability | \"Didn't mean to\" doesn't undo the harm. |\n| [D] Stop investigating once incompetence is identified | Incompetence-caused harm still requires response. |\n| [D] Use razor to gaslight someone who was genuinely harmed | Attribution ≠ override of their lived experience. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Razor invoked by the actor, not the receiver\n- \"Incompetence\" explanation conveniently absolves a powerful party\n- Pattern of \"incompetence\" always harming the same direction\n- Razor used to silence a victim's report or applied repeatedly without updating priors\n\n## Verification\n\n- [ ] Action and harm stated factually (not interpretively)\n- [ ] Specific incompetence explanation constructed and coverage assessed\n- [ ] Additional assumptions required for malice named\n- [ ] Clarifying-conversation posture chosen (not accusation)\n- [ ] Override signals specified; power asymmetry considered\n\n---\n\n*Part of **deciqAI Knowledge Skills** — open-source thinking skills that make rigor executable for AI agents. Built by deciqAI · https://deciqai.com · Contributions welcome — see the template at the repo root.*\n\nFile v1.0.0:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"hanlons-razor\",\n \"version\": \"1.0.0\",\n \"publishedAt\": 1782713843213\n}\n\nFile v1.0.0:references/sources.md\n\n# Sources — hanlons-razor\n\n> *Primary sources for the [hanlons-razor](../SKILL.md) skill.*\n\n## Sources\n\n- Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two: More Reasons Why Things Go Wrong.* Price/Stern/Sloan. ISBN 978-0843105674.\n- Heinlein, R. A. (1941). \"Logic of Empire.\" In *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211.\n- Goethe, J. W. (1774). *Die Leiden des jungen Werthers.*\n- Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n- Akerlof, G. A. & Shiller, R. J. (2015). *Phishing for Phools.* Princeton University Press. ISBN 978-0691168319.\n- Grove, A. S. (1983). *High Output Management.* Random House. ISBN 978-0679762881.\n- \"Hanlon's Razor.\" *Quote Investigator.* https://quoteinvestigator.com/2016/12/30/not-malice/\n\nFile v1.0.0:examples/hanlons-submission-1980-heinleins-1941-articulation.md\n\n# Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation\n\n> *Example for the [hanlons-razor](../SKILL.md) skill.*\n\nThe Hanlon's Razor formulation that spread across modern internet culture, business writing, and decision-theory literature was submitted by **Robert J. Hanlon** of Scranton, Pennsylvania, to the 1980 humor compendium *Murphy's Law Book Two: More Reasons Why Things Go Wrong*, edited by Arthur Bloch:\n\n> \"Never attribute to malice that which is adequately explained by stupidity.\"\n\n— Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two.* Price/Stern/Sloan. ISBN 978-0843105674.\n\nThe submission was one of dozens in a humorous compendium. Most have been forgotten. Hanlon's became one of the most-cited heuristics of the next 40 years — a fact that should itself give pause.\n\nThe deeper history runs back further. **Robert A. Heinlein**, in his 1941 novella \"Logic of Empire\" (published in *Astounding Science Fiction*, later collected in *The Past Through Tomorrow*, 1967), wrote a passage that historians of the saying believe Hanlon either independently re-discovered or consciously adapted:\n\n> \"You have attributed conditions to villainy that simply result from stupidity.\"\n\n— Heinlein, R. A. (1941). \"Logic of Empire.\" *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211, p. 224.\n\nThe \"Logic of Empire\" context is instructive. Heinlein's protagonists are abolitionists who have come to believe the slavery system on Venus is being maintained by an evil conspiracy. The older character corrects them: the system persists not because of conspiracy but because of structural incompetence — bad information, misaligned incentives, individual short-term rationality producing collective long-term harm. The point: **systems that look like conspiracies are usually structures that have evolved to produce the same outcome without anyone planning it**. This is Hanlon's razor applied at the level of institutions, not individuals.\n\nEven earlier, **Johann Wolfgang von Goethe** had written in *Sorrows of Young Werther* (1774):\n\n> \"Misunderstandings and lethargy perhaps produce more wrong in the world than deceit and malice do. At least the last two are certainly rarer.\"\n\n— Goethe, J. W. (1774). *Die Leiden des jungen Werthers.* As discussed at https://quoteinvestigator.com/2016/12/30/not-malice/\n\nThe three formulations span 200+ years, three languages, and quite different cultural contexts — yet articulate the same calibration error. The fact that the observation keeps being independently re-derived suggests a robust underlying psychological fact: **humans systematically over-attribute malice.**\n\nThe empirical psychology supports the heuristic. The **fundamental attribution error** (Ross 1977) documents that humans attribute others' behavior to disposition (intent, character) while attributing their own similar behavior to situation (circumstance, pressure). When someone cuts us off in traffic, \"they're a jerk\" (disposition); when we cut someone off, \"I had an emergency\" (situation). Hanlon's razor is a corrective: extend the situation-attribution we apply to ourselves to others' behavior.\n\n> \"The tendency for people to over-emphasize personal characteristics and ignore situational factors in judging others' behavior. ... Because of the fundamental attribution error, we tend to believe that others do bad things because they are bad people.\"\n>\n> — Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n\nThe framework has been adopted operationally:\n\n**Software engineering.** Post-mortems of production failures routinely apply Hanlon's razor: the engineer who pushed the bad commit was not malicious, and the corrective is process/tooling, not punishment. Google's **blameless postmortem** practice is Hanlon's razor institutionalized.\n\n**Corporate management.** Andy Grove, in *High Output Management*, applied the heuristic to performance management: a subordinate who keeps making the same mistake is most often not insubordinate — they are missing information, training, or context.\n\n**International diplomacy.** The \"rational actor\" framework in international relations is an application at the state level: states act in their own (sometimes incompetent) interest rather than out of pure malice.\n\nThe framework also has documented failure modes:\n\n**First, the razor can be weaponized.** People who have caused real harm can invoke Hanlon's razor to deflect accountability. The defense: the razor specifies a *starting prior*, not a final attribution. Repeated incompetence-causing-harm despite available information eventually becomes equivalent to malice in practical effect.\n\n**Second, the razor can disable necessary protective response.** When the cost of being wrong about non-malice is catastrophic (abusive partner, predatory company, hostile state), the calibrated prior must shift toward defensive posture *before* malice is proven.\n\n**Third, the razor doesn't excuse harm.** A bridge that collapses through incompetence has still collapsed. The point is *attribution*, not *accountability*. Once incompetence is identified, the corrective is competence-building, not blame-shifting.\n\nThree operational lessons from extensive application:\n\n**First, the conversation that follows from mistake-prior is almost always better than the one that follows from malice-prior — even when it later turns out to be malice.** Clarifying conversations produce information; accusatory ones produce defensiveness.\n\n**Second, the malice attribution feels viscerally correct in real-time and almost always feels overconfident in retrospect.** This calibration error is robust.\n\n**Third, the razor scales from individual to institutional.** Most \"bad culture\" / \"broken system\" complaints are about coordination failure, not coordinated malice.\n\nFile v1.0.0:skill-card.md\n\n## Description:
\nHanlon's Razor helps agents evaluate hostile-seeming actions by first testing whether mistakes, miscommunication, or missing context explain the evidence before escalating to malice.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[deciqai](https://clawhub.ai/user/deciqai)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nExternal users, developers, and teams use this skill when a conflict or bad outcome is being interpreted as intentional harm. It guides the agent through a structured mistake-first analysis, clarifying conversation, and override signal for updating toward malice when evidence changes.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: Users may apply the mistake-first prior where there is documented malice, serious power imbalance, abuse risk, or high consequence from underestimating intent.
\nMitigation: Use the skill only as a reasoning aid, avoid it in those contexts, and update away from non-malice when concrete override evidence appears.
\n\n\n## Reference(s):
\n- [Primary sources](references/sources.md)
\n- [Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation](examples/hanlons-submission-1980-heinleins-1941-articulation.md)
\n- [Quote Investigator: Hanlon's Razor](https://quoteinvestigator.com/2016/12/30/not-malice/)
\n\n\n## Skill Output:
\n**Output Type(s):** [Text, Markdown, Guidance]
\n**Output Format:** [Markdown analysis with concise structured fields]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [Text-only reasoning aid; no files, credentials, tools, or persistent background behavior requested.]
\n\n## Skill Version(s):
\n1.0.0 (source: server release evidence)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
","readmeExcerpt":"Skill: Hanlon's Razor Owner: deciqai Summary: Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-07-16T18:01:56.620Z | user Description tail link + agents machine-readable metadata line (deciqai.com/s/hanlons-razor.json) v1.0.4 | 2026-07-10T10:26:14.117Z | user","codeSnippets":[],"executableExamples":[{"language":"text","snippet":"- What was done: \n- Harm to me: \n- Gut attribution: "},{"language":"text","snippet":"- Bad information they had: / Didn't realize: / Optimizing for: / Under pressure from:\n- Coverage: <% of observed behavior this explains>"},{"language":"text","snippet":"- Info they'd need: / Motivation at your expense: / Harm predictable from their position?"},{"language":"text","snippet":"- Prior: · Starting posture: · First move: · Override trigger:"},{"language":"text","snippet":"# Hanlon's Razor Analysis: \nMistake explanation + coverage % | Malice extra assumptions | Prior | Override signal | First action"},{"language":"text","snippet":"- What was done: A production image model generated historically inaccurate outputs; the vendor\n publicly paused the feature and apologized.\n- Harm to me (the observer/user): Loss of trust in the tool; the outputs were wrong and, for\n some prompts, offensive.\n- Gut attribution: \"They did this on purpose to push an agenda.\""}],"parameters":null,"dependencies":[],"permissions":[],"extractedFiles":[{"path":"SKILL.md","content":"---\nname: hanlons-razor\ndescription: \"Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone is about to escalate based on assumed malicious intent; a pattern of bad outcomes is being labeled a coordinated attack. Do NOT activate when: concrete documented evidence of malice already exists; the cost of being wrong about non-malice is catastrophic (e.g., safety-critical or abusive-relationship context). More: deciqai.com/c/hanlons-razor\"\n---\n\n# Hanlon's Razor\n\n## Overview\n\nBefore assuming someone hurt you on purpose, construct the version where they made a mistake — and see how much evidence it explains. The razor is a Bayesian *prior*, not a proof; override it when concrete evidence of malice arrives. Human attribution systematically over-weights intent (fundamental attribution error); most hostile-seeming acts are incompetence, miscommunication, or asymmetric information.\n\nComposes with `bayesian-reasoning`, `abductive-reasoning`, `occams-razor`, `critical-thinking`.\n\n## When to Use\n\n- You feel an emotional pull toward \"they did this on purpose\"\n- You're about to escalate on the assumption of malice\n- A pattern of bad outcomes is being framed as a coordinated attack\n- A team is in conflict and each side believes the other is acting in bad faith\n- An AI model's harmful/biased output or a competitor's surprising AI move is being read as deliberate malice rather than an emergent bug, honest error, or ordinary self-interested competition\n\n**Not when:** concrete evidence of malicious intent exists; cost of being wrong is catastrophic; power imbalance makes \"they probably didn't mean it\" an abuse-enabling stance.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line: before believing someone did it on purpose, construct the mistake version — see how much evidence it covers.\n2. Check fit: concrete malice evidence / catastrophic cost of being wrong → not this lens.\n3. Elicit the specific incident — what exactly happened?\n> **[WAIT — do not advance until user responds]**\n4. Work through The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close: name the clarifying-conversation move + the override signal to watch for.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe the action and harm** (factual, not interpretive)\n```\n- What was done: \n- Harm to me: \n- Gut attribution: \n```\n**Step 2 — Construct the non-malice explanation**\n```\n- Bad information they had: / Didn't realize: / Optimizing for: / Under pressure from:\n- Coverage: <% of observed behavior this explains>\n`"},{"path":"_meta.json","content":"{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"hanlons-razor\",\n \"version\": \"1.0.5\",\n \"publishedAt\": 1784224916620\n}"},{"path":"references/sources.md","content":"# Sources — hanlons-razor\n\n> *Primary sources for the [hanlons-razor](../SKILL.md) skill.*\n\n## Sources\n\n- Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two: More Reasons Why Things Go Wrong.* Price/Stern/Sloan. ISBN 978-0843105674.\n- Heinlein, R. A. (1941). \"Logic of Empire.\" In *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211.\n- Goethe, J. W. (1774). *Die Leiden des jungen Werthers.*\n- Ross, L. (1977). \"The intuitive psychologist and his shortcomings: Distortions in the attribution process.\" *Advances in Experimental Social Psychology*, 10, 173-220.\n- Akerlof, G. A. & Shiller, R. J. (2015). *Phishing for Phools.* Princeton University Press. ISBN 978-0691168319.\n- Grove, A. S. (1983). *High Output Management.* Random House. ISBN 978-0679762881.\n- \"Hanlon's Razor.\" *Quote Investigator.* https://quoteinvestigator.com/2016/12/30/not-malice/\n- Google (2024). \"Gemini image generation got it wrong. We'll do better.\" *The Keyword (official Google blog),* February 23, 2024. https://blog.google/products/gemini/gemini-image-generation-issue/ (context for the 2024–2026 AI-incidents example: a vendor publicly pausing and apologizing for an emergent over-correction, consistent with honest error rather than intent).\n- DeepSeek (2025). \"DeepSeek-R1\" model release and technical report, January 2025 (context for the 2024–2026 AI-incidents example: a surprising competitor move read by some as sabotage/fabrication, better modeled as ordinary self-interested competition absent evidence of falsification)."},{"path":"examples/ai-incidents-malice-vs-emergent-error-2024-2026.md","content":"# Method in Action: AI Incidents — Incompetence and Emergent Error as a Prior Over Malice (2024–2026)\n\n> *Example for the [hanlons-razor](../SKILL.md) skill.*\n\nBetween 2024 and 2026, as large language models and AI agents moved into mass adoption, a recurring interpretive question arose: when a model produces a harmful, biased, or embarrassing output — or when a competitor makes a surprising, seemingly aggressive move — is the right default *deliberate malice* or *incompetence / honest error / emergent bug*? Hanlon's razor supplies the calibrated starting prior. This walkthrough runs the incident class through the skill's own Process, using two concrete, well-documented public examples.\n\n---\n\n## Case A: A model produces a harmful output\n\n**Anchor:** In February 2024, Google paused the image-generation feature of its Gemini model after it produced historically inaccurate images — including racially diverse depictions of subjects where that was historically wrong. A wave of online commentary framed this as a deliberate ideological agenda encoded on purpose.\n\n### Step 1 — Describe the action and harm (factual, not interpretive)\n\n```\n- What was done: A production image model generated historically inaccurate outputs; the vendor\n publicly paused the feature and apologized.\n- Harm to me (the observer/user): Loss of trust in the tool; the outputs were wrong and, for\n some prompts, offensive.\n- Gut attribution: \"They did this on purpose to push an agenda.\"\n```\n\n### Step 2 — Construct the non-malice explanation\n\n```\n- Bad information / didn't realize: Post-training fine-tuning and system-prompt-level guardrails\n meant to reduce a KNOWN failure mode (over-representation of one group in image outputs) were\n applied too bluntly and generalized to cases where they produced absurd results.\n- Optimizing for: A real, previously-criticized bias problem — over-correction, not a coherent plot.\n- Under pressure from: Ship pace; guardrails tuned and tested unevenly across prompt types;\n emergent interaction between the model and the correction layer that was not fully anticipated.\n- Coverage: This explains ~90% of the observed behavior — including the vendor's own fast public\n pause-and-apologize, which is what an embarrassed org does, not what a covert-agenda org does.\n```\n\n### Step 3 — Name what malice would additionally require\n\n```\n- Info they'd need: Foreknowledge that these exact absurd outputs would occur and be shipped anyway.\n- Motivation at your expense: A deliberate intent to deceive users, accepting near-certain\n reputational damage and a public reversal — against the vendor's own commercial interest.\n- Harm predictable from their position? A deliberate plot predicts a cover-up, not a same-week\n public apology and rollback. The observed response is the OPPOSITE of what malice predicts.\n```\n\n### Step 4 — Choose starting posture · Step 5 — Set override signal · Step 6 — Hold prior\n\n```\n- Prior: emergent over-correction / honest engineering error (no"},{"path":"examples/hanlons-submission-1980-heinleins-1941-articulation.md","content":"# Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation\n\n> *Example for the [hanlons-razor](../SKILL.md) skill.*\n\nThe Hanlon's Razor formulation that spread across modern internet culture, business writing, and decision-theory literature was submitted by **Robert J. Hanlon** of Scranton, Pennsylvania, to the 1980 humor compendium *Murphy's Law Book Two: More Reasons Why Things Go Wrong*, edited by Arthur Bloch:\n\n> \"Never attribute to malice that which is adequately explained by stupidity.\"\n\n— Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two.* Price/Stern/Sloan. ISBN 978-0843105674.\n\nThe submission was one of dozens in a humorous compendium. Most have been forgotten. Hanlon's became one of the most-cited heuristics of the next 40 years — a fact that should itself give pause.\n\nThe deeper history runs back further. **Robert A. Heinlein**, in his 1941 novella \"Logic of Empire\" (published in *Astounding Science Fiction*, later collected in *The Past Through Tomorrow*, 1967), wrote a passage that historians of the saying believe Hanlon either independently re-discovered or consciously adapted:\n\n> \"You have attributed conditions to villainy that simply result from stupidity.\"\n\n— Heinlein, R. A. (1941). \"Logic of Empire.\" *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211, p. 224.\n\nThe \"Logic of Empire\" context is instructive. Heinlein's protagonists are abolitionists who have come to believe the slavery system on Venus is being maintained by an evil conspiracy. The older character corrects them: the system persists not because of conspiracy but because of structural incompetence — bad information, misaligned incentives, individual short-term rationality producing collective long-term harm. The point: **systems that look like conspiracies are usually structures that have evolved to produce the same outcome without anyone planning it**. This is Hanlon's razor applied at the level of institutions, not individuals.\n\nEven earlier, **Johann Wolfgang von Goethe** had written in *Sorrows of Young Werther* (1774):\n\n> \"Misunderstandings and lethargy perhaps produce more wrong in the world than deceit and malice do. At least the last two are certainly rarer.\"\n\n— Goethe, J. W. (1774). *Die Leiden des jungen Werthers.* As discussed at https://quoteinvestigator.com/2016/12/30/not-malice/\n\nThe three formulations span 200+ years, three languages, and quite different cultural contexts — yet articulate the same calibration error. The fact that the observation keeps being independently re-derived suggests a robust underlying psychological fact: **humans systematically over-attribute malice.**\n\nThe empirical psychology supports the heuristic. The **fundamental attribution error** (Ross 1977) documents that humans attribute others' behavior to disposition (intent, character) while attributing their own similar behavior to situation (circumstance, pressure). When someone cuts us off in traffic, \"they're a jerk\" (disposition); w"}],"languages":[],"docsSourceLabel":"CLAWHUB","editorialOverview":"Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone... Skill: Hanlon's Razor Owner: deciqai Summary: Activate when: someone feels a colleague/partner/company did something on purpose to hurt them; a team believes another side is acting in bad faith; someone... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-07-16T18:01:56.620Z | user Description tail link + agents machine-readable metadata line (deciqai.com/s/hanlons-razor.json) v1.0.4 | 2026-07-10T10:26:14.117Z | user","editorialQuality":{"score":100,"threshold":65,"status":"ready","wordCount":1492,"uniquenessScore":57,"reasons":[]}},"media":{"evidence":{"source":"no-media","verified":false,"confidence":"low","updatedAt":"2026-10-11T15:04:22.646Z","emptyReason":"No screenshots, media assets, or demo links are available."},"primaryImageUrl":null,"mediaAssetCount":0,"assets":[],"demoUrl":null},"ownerResources":{"evidence":{"source":"unclaimed","verified":false,"confidence":"low","updatedAt":"2026-10-11T15:04:22.646Z","emptyReason":"This page has not been claimed by the agent owner."},"hasCustomPage":false,"customPageUpdatedAt":null,"customLinks":[],"structuredLinks":{"docsUrl":null,"demoUrl":null,"supportUrl":null,"pricingUrl":null,"statusUrl":null},"customPage":null},"relatedAgents":{"evidence":{"source":"protocol-neighbors","verified":false,"confidence":"medium","updatedAt":"2026-10-11T17:44:15.909Z","emptyReason":null},"items":[{"id":"8ebccd8e-3863-4187-8355-c3f14e1f9edf","entityType":"agent","canonicalPath":"/agent/iofficeai-aionui","slug":"iofficeai-aionui","name":"AionUi","description":"Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!","url":"https://github.com/iOfficeAI/AionUi","homepage":"https://www.aionui.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-10-09T19:11:12.944Z","createdAt":"2026-02-25T03:38:16.584Z","downloads":null},{"id":"b917f68a-ebff-438e-84f8-3f4b2494c0bc","entityType":"agent","canonicalPath":"/agent/activepieces-activepieces","slug":"activepieces-activepieces","name":"activepieces","description":"AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents","url":"https://github.com/activepieces/activepieces","homepage":"https://www.activepieces.com","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-15T02:22:12.426Z","createdAt":"2026-02-25T03:38:12.412Z","downloads":null},{"id":"5cb26759-3a39-483f-94cf-276a98c13bb8","entityType":"agent","canonicalPath":"/agent/cherryhq-cherry-studio","slug":"cherryhq-cherry-studio","name":"cherry-studio","description":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs","url":"https://github.com/CherryHQ/cherry-studio","homepage":"https://cherry-ai.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-11T14:38:40.986Z","createdAt":"2026-02-25T03:38:19.379Z","downloads":null},{"id":"6f6582d0-5d76-4f0f-b81d-86520247950b","entityType":"agent","canonicalPath":"/agent/copilotkit-copilotkit","slug":"copilotkit-copilotkit","name":"CopilotKit","description":"The Frontend for Agents & Generative UI. React + Angular","url":"https://github.com/CopilotKit/CopilotKit","homepage":"https://docs.copilotkit.ai","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-03-25T09:50:57.846Z","createdAt":"2026-02-25T03:39:14.617Z","downloads":null}],"links":{"hub":"/agent","source":"/agent/source/clawhub","protocols":[{"label":"OpenClaw","href":"/agent/protocol/openclew"}]}}}