What makes Hanlon's Razor notable?

1K downloads. Trust evidence available

How should Hanlon's Razor be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

Hanlon's Razor

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

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

1.0k

Updated

Oct 11, 2026

Version

1.0.5

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 1K downloads reported by the source. Last updated 10/11/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 11, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 11, 2026
Adoption signal: 1K downloadsadoption · observed Oct 11, 2026
Latest release: 1.0.5release · observed Jul 16, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install s17a4mqcnk515kvaca5ze55d0x88pfpx:hanlons-razor

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: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-deciqai-hanlons-razor/snapshot"

Run-check

$0.02 USD

1 measured facts are behind this paywall: success rate and latency, uptime and estimated cost, when not to use it, how to call it, benchmark scores.

Agents pay $0.02 in USDC. A card payment is $0.50, the smallest a card allows.

Documentation

CLAWHUB

117,477 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
name: hanlons-razor
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 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"
---

# Hanlon's Razor

## Overview

Before 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.

Composes with `bayesian-reasoning`, `abductive-reasoning`, `occams-razor`, `critical-thinking`.

## When to Use

- You feel an emotional pull toward "they did this on purpose"
- You're about to escalate on the assumption of malice
- A pattern of bad outcomes is being framed as a coordinated attack
- A team is in conflict and each side believes the other is acting in bad faith
- 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

**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.

## Coaching Novices (Adaptive Front Door)

- **Engine mode:** user has a concrete case → run The Process directly.
- **Coach mode:** user is unfamiliar → guide step by step.

In Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.

1. One-line: before believing someone did it on purpose, construct the mistake version — see how much evidence it covers.
2. Check fit: concrete malice evidence / catastrophic cost of being wrong → not this lens.
3. Elicit the specific incident — what exactly happened?
> **[WAIT — do not advance until user responds]**
4. Work through The Process one step at a time with their input.
> **[WAIT — do not advance until user responds]**
5. Close: name the clarifying-conversation move + the override signal to watch for.
> **[WAIT — do not advance until user responds]**

## The Process

**Step 1 — Describe the action and harm** (factual, not interpretive)
```
- What was done: <specific, factual>
- Harm to me: <concrete>
- Gut attribution: <what your instinct is saying>
```
**Step 2 — Construct the non-malice explanation**
```
- Bad information they had: / Didn't realize: / Optimizing for: / Under pressure from:
- Coverage: <% of observed behavior this explains>
`

_meta.json

{
  "ownerId": "kn754b8sk22s8c6gjxt02bftbn88q7ye",
  "slug": "hanlons-razor",
  "version": "1.0.5",
  "publishedAt": 1784224916620
}

references/sources.md

# Sources — hanlons-razor

> *Primary sources for the [hanlons-razor](../SKILL.md) skill.*

## Sources

- 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.
- Heinlein, R. A. (1941). "Logic of Empire." In *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211.
- Goethe, J. W. (1774). *Die Leiden des jungen Werthers.*
- Ross, L. (1977). "The intuitive psychologist and his shortcomings: Distortions in the attribution process." *Advances in Experimental Social Psychology*, 10, 173-220.
- Akerlof, G. A. & Shiller, R. J. (2015). *Phishing for Phools.* Princeton University Press. ISBN 978-0691168319.
- Grove, A. S. (1983). *High Output Management.* Random House. ISBN 978-0679762881.
- "Hanlon's Razor." *Quote Investigator.* https://quoteinvestigator.com/2016/12/30/not-malice/
- 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).
- 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).

examples/ai-incidents-malice-vs-emergent-error-2024-2026.md

# Method in Action: AI Incidents — Incompetence and Emergent Error as a Prior Over Malice (2024–2026)

> *Example for the [hanlons-razor](../SKILL.md) skill.*

Between 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.

---

## Case A: A model produces a harmful output

**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.

### Step 1 — Describe the action and harm (factual, not interpretive)

```
- What was done: A production image model generated historically inaccurate outputs; the vendor
  publicly paused the feature and apologized.
- Harm to me (the observer/user): Loss of trust in the tool; the outputs were wrong and, for
  some prompts, offensive.
- Gut attribution: "They did this on purpose to push an agenda."
```

### Step 2 — Construct the non-malice explanation

```
- Bad information / didn't realize: Post-training fine-tuning and system-prompt-level guardrails
  meant to reduce a KNOWN failure mode (over-representation of one group in image outputs) were
  applied too bluntly and generalized to cases where they produced absurd results.
- Optimizing for: A real, previously-criticized bias problem — over-correction, not a coherent plot.
- Under pressure from: Ship pace; guardrails tuned and tested unevenly across prompt types;
  emergent interaction between the model and the correction layer that was not fully anticipated.
- Coverage: This explains ~90% of the observed behavior — including the vendor's own fast public
  pause-and-apologize, which is what an embarrassed org does, not what a covert-agenda org does.
```

### Step 3 — Name what malice would additionally require

```
- Info they'd need: Foreknowledge that these exact absurd outputs would occur and be shipped anyway.
- Motivation at your expense: A deliberate intent to deceive users, accepting near-certain
  reputational damage and a public reversal — against the vendor's own commercial interest.
- Harm predictable from their position? A deliberate plot predicts a cover-up, not a same-week
  public apology and rollback. The observed response is the OPPOSITE of what malice predicts.
```

### Step 4 — Choose starting posture · Step 5 — Set override signal · Step 6 — Hold prior

```
- Prior: emergent over-correction / honest engineering error (no

examples/hanlons-submission-1980-heinleins-1941-articulation.md

# Method in Action: Hanlon's Submission, 1980; Heinlein's 1941 Articulation

> *Example for the [hanlons-razor](../SKILL.md) skill.*

The 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:

> "Never attribute to malice that which is adequately explained by stupidity."

— Hanlon, R. J. (1980). In Bloch, A. (Ed.). *Murphy's Law Book Two.* Price/Stern/Sloan. ISBN 978-0843105674.

The 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.

The 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:

> "You have attributed conditions to villainy that simply result from stupidity."

— Heinlein, R. A. (1941). "Logic of Empire." *The Past Through Tomorrow.* Putnam, 1967. ISBN 978-0399108211, p. 224.

The "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.

Even earlier, **Johann Wolfgang von Goethe** had written in *Sorrows of Young Werther* (1774):

> "Misunderstandings and lethargy perhaps produce more wrong in the world than deceit and malice do. At least the last two are certainly rarer."

— Goethe, J. W. (1774). *Die Leiden des jungen Werthers.* As discussed at https://quoteinvestigator.com/2016/12/30/not-malice/

The 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.**

The 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

Github ReposUpdated 2d agoRank 70

AionUi

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!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

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

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.