What it is
discernment-nudge is a skill that runs before finalizing a substantive reply the user may act on, such as advice, drafted plans, estimates, data analysis, or factual claims. If it applies, it appends two or three short follow-up questions tied to specifics in the answer. The questions model three AI Fluency discernment habits: checking facts, questioning reasoning, and noticing missing context. It fires at most once per conversation and has explicit exclusions for trivial, creative, or already-verified requests.
Who it's for
- Skill authors who want Claude to prompt users to scrutinize consequential answers (money, health, legal, career, plans)
- Users who tend to take confident, well-structured AI answers at face value
- Teams studying how to model AI Fluency discernment habits in assistant behavior
Requirements
Requirements
- A Claude environment that supports skills, with the SKILL.md file available
- License terms are in LICENSE.txt (per the skill frontmatter)
Examples
Exact nudge output format
PromptA few things worth a second look:
- How do these CPL estimates compare to benchmarks in my specific vertical?
- Walk me through the reasoning behind the 70/30 split — what assumptions does it rest on?Expected output: The nudge is plain text appended after a blank line at the end of the answer, using the exact lead-in line followed by plain bullets, with nothing after it.
Fact-or-figure prompt
PromptHow do these CPL estimates compare to benchmarks in my specific vertical?Expected output: Points at a specific figure in the answer and asks how to check it or compare it to the user's own data.
Reasoning-step prompt
PromptWalk me through why you prioritized webinars over content — what assumptions does that rest on?Expected output: Points at a reasoning step and invites the user to probe the assumptions behind it.
Missing-context prompt
PromptI didn't mention my state — does the security-deposit rule change by jurisdiction?Expected output: Points at context the answer had to guess at. Prompts are phrased in first person, as something the user could send back verbatim, under ~120 characters.
Pros & cons
Pros
- Pro:Prompts must reference concrete items (a number, step, or assumption) in the answer, avoiding generic 'verify this' filler
- Pro:Clear carve-outs (creative writing, casual chat, code to be run, simple lookups, educational explanations) reduce noise
- Pro:Respects user intent: skips the nudge when the user asked for verification, a quick version, a review, or gave their own material
- Pro:Limited to once per conversation and 2-3 short bullets, so it stays light and easy to skip
Cons
- Con:Judging whether an answer is 'actable' versus purely educational involves boundary calls that may be inconsistent
- Con:Strict output format (fixed lead-in, plain text, nothing after) leaves little room for customization
- Con:Fires at most once per conversation, so later qualifying answers get no nudge
Images
Related
Skill 
Skill algorithmic-art
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
Skill brand-guidelines
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
Skill canvas-design
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.
Skill claude-api
|-
Skill doc-coauthoring
Guide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.