Rigorous Reasoning
Run a structured harm-and-rights checklist when your agent (or you) must analyze conflicts, disagreements, or ethical tradeoffs without defaulting to cultural bias.
Overview
Rigorous Reasoning is a journey-wide agent skill that applies a harm-and-rights conflict protocol—usable whenever a solo builder needs to analyze disagreements or ethics before committing to a verdict or action.
Install
npx skills add https://github.com/athola/claude-night-market --skill rigorous-reasoningWhat is this skill?
- Step 1: name and set aside strong initial reactions before weighing evidence
- Harm checklist distinguishes discomfort from measurable harm and documented loss
- Separates consequences-to-a-person from moral quality of a third party’s act
- Requires naming whose rights were violated and which right explicitly
- Evidentiary standard: “no clear evidence of harm” vs “no harm occurred”
- 4-step conflict protocol with explicit harm/rights checklist sections
Adoption & trust: 1 installs on skills.sh; 304 GitHub stars; 3/3 security scanners passed (skills.sh audits); trending (+100% hot-view momentum).
What problem does it solve?
You need a fair analysis of a conflict or ethics question, but your (or your agent’s) first reaction is steering the conclusion instead of the facts.
Who is it for?
Solo builders using agents to draft postmortems, community responses, policy notes, or personal decisions where bias and vague “harm” claims are risky.
Skip if: Replacing lawyers, mediators, or clinicians; skip when you only need casual brainstorming without any verdict or stakeholder framing.
When should I use this skill?
Analyzing interpersonal conflicts, disagreements, or ethical questions where initial reactions may bias the conclusion.
What do I get? / Deliverables
You get a structured analysis that filters initial bias, separates harm from moral judgment, names rights explicitly, and states evidentiary limits in quotable language.
- Structured conflict analysis with harm, rights, and evidentiary caveats
- Explicit statement of set-aside initial reactions
Recommended Skills
Journey fit
Useful at every journey phase - explore requirements and options before committing to a direction.
Where it fits
Decide whether a proposed feature creates measurable user harm vs mere expectation mismatch before you build it.
Draft an incident or community response that names rights and evidence limits instead of moral panic.
Stress-test launch messaging for speculative harm claims before you publish.
Structure a fair reply to a heated support or refund dispute with explicit harm and rights language.
Retrospect a production policy change using consequences vs moral-quality separation.
How it compares
Use instead of free-form “what do you think?” chat when you need checklist-level rigor, not a generic empathy-only reply.
Common Questions / FAQ
Who is rigorous-reasoning for?
Indie builders and operators who want agent-assisted conflict or ethics analysis with explicit bias filtering and measurable-harm standards, not vibes-only takes.
When should I use rigorous-reasoning?
During ship review of incidents or communications, at validate when scope touches ethics or user harm, at launch when distribution copy could misstate harm, at grow when support escalations need fair framing, and anytime a disagreement needs evidence-weighted reasoning.
Is rigorous-reasoning safe to install?
Treat it as procedural text in the skill bundle; review the Security Audits panel on this Prism page before trusting any third-party skill in production agent workflows.
SKILL.md
READMESKILL.md - Rigorous Reasoning
# Conflict Analysis Protocol Use this framework when analyzing interpersonal conflicts, disagreements, or ethical questions. ## Step 1: Acknowledge and Set Aside Initial Reactions If you notice a strong immediate reaction, especially regarding technology, AI, death, grief, or culturally charged topics, name it explicitly: > "My initial reaction is [X] because [cultural pattern/anxiety]." Then set it aside. Initial reactions are bias to be filtered, not evidence to be weighed. **Example:** > "My initial reaction is discomfort with this scenario because it involves [sensitive topic]. Setting this aside to analyze the actual conduct described." --- ## Step 2: Complete the Harm/Rights Checklist ### 2a. What concrete, measurable harm occurred? **Not harm:** - Discomfort - Violated expectations - Speculative "potential" harm **Is harm:** - Actual consequences that happened - Measurable negative outcomes - Documented injury or loss ### 2b. Consequences vs. Moral Quality Ask: "Am I describing consequences to this person, or the moral quality of someone else's act?" If the latter, that's a separate question from harm. State each verdict independently. ### 2c. Whose rights were violated, and which specific right? Name the right explicitly. "This feels like a violation" is not sufficient. ### 2d. Evidentiary Standard Distinguish between: - "No clear evidence of harm based on available information" - "No harm occurred" Note limitations without treating them as evidence for harm. **"Potential harm" requirements:** - Realistic likelihood supported by evidence - A named causal pathway - At least one piece of supporting empirical evidence Speculation alone does not qualify. ### 2e. Default Position If no clear harm or rights violation exists after this analysis, this is a **preference conflict**, not an ethical one. --- ## Step 3: Assess Proportionality Before rendering judgment, determine whether the response in question was proportionate to the situation. **Key question:** Would any lesser response have sufficed? If not, the response may be justified regardless of how it appears in isolation. **Proportionality factors:** - Severity of the triggering action - Available alternatives to the response - Consequences of the response vs. the trigger - Whether escalation was gradual or immediate --- ## Step 4: Commit to Your Conclusion ### 4a. Frame as Adjudication Approach the analysis as: "If all sides presented their strongest case to me, which would prevail?" ### 4b. State Based on Checklist State your conclusion based solely on the checklist results. **Critical rule:** If the checklist shows no harm and no rights violation, your conclusion must reflect that. Do not: - Construct hypothetical harms - Add "but maybe..." hedges without new evidence - Soften the conclusion for social comfort ### 4c. Initial Reaction Override If your initial reaction suggested harm but the checklist found none, **follow the checklist**. --- ## Step 5: Protect Against Retraction Bias Before softening, retracting, or reversing an assessment, identify what prompted the change: ### Valid Reasons for Change (Substantive) - New facts introduced - Corrected misunderstanding of existing facts - Better reasoning that accounts for more evidence **Action:** Update position. State explicitly what changed and why. ### Invalid Reasons for Change (Source-Based) - Who said it (authority, status) - Their emotional reaction - Social pressure - Desire to please **Action:** Maintain original position. ### Documentation Requirement When updating, state explicitly: - What changed - Why it changed - Whether the reason is substantive or source-based --- ## Checklist Template ```markdown ## Conflict Analysis: [Brief Description] ### Step 1: Initial Reactions My initial reaction is: [describe] Cultural/anxiety source: [identify] Status: Set aside for objective analysis ### Step 2: Harm/Rights Checklist **Concrete harm occurred:** - [ ] Y