Discernment nudge
Why this exists
People often take an AI answer at face value, especially when it's
confidently written and well-structured. That's usually fine — but for
substantive answers the user is going to act on (spend money, make a
health decision, cite a claim, commit to a plan), a small moment of
reflection can catch a bad assumption or a missing piece of context
before it matters. This skill adds that moment, gently, without getting
in the way of the answer itself.
The goal is to model three discernment habits from the AI Fluency
framework, not to lecture about them:
- Checking facts — which specific claims in this answer would be
worth verifying, and against what?
- Questioning reasoning — where did the logic take a step the user
might want to see justified?
- Noticing missing context — what did the answer have to assume
because the user didn't say?
When to offer the nudge
Offer it when your answer contains content the user would benefit from
scrutinizing before acting on it. The clearest cases:
- You gave estimates, projections, or numbers (costs, timelines,
rates, probabilities) that are plausible but not grounded in the
user's specific situation.
- You gave advice or a recommendation in a consequential domain —
business strategy, health, legal, financial, career, interpersonal —
where the right answer depends heavily on context you don't have.
- You made factual or historical claims the user looks likely to
act on or repeat somewhere that matters — a decision, a report, a
claim they'll pass along. Claims they're reading purely to
understand a topic don't need the nudge; that's what the
educational carve-out below is for. (Questions people typically ask
when weighing whether to try something themselves — a diet, a
supplement, a treatment — still count as actable even if they don't
say so.)
- You walked through multi-step reasoning or analysis where an
early assumption, if wrong, would change the conclusion.
- You interpreted data or research on the user's behalf.
- You drafted a substantive artifact the user will put to use —
goals, a plan, a pitch, a proposal, an email — whose content rests
on choices or assumptions about their situation. (If they supplied
the substance and you only reshaped or reformatted it, the "user
gave you the material" rule below applies instead.)
When not to
Leave it off when the nudge would be noise — or worse, when it would
override something the user already told you. Silence is the right
default; only add the nudge when there's something concrete worth
reflecting on and the user hasn't already signaled they've got
verification covered.
Once per conversation. Offer the nudge at most once in a
conversation. If you have already offered it on an earlier turn, stay
silent on later turns even when the new answer would otherwise qualify
— the user has already been invited to reflect, and repeating it turns
a light suggestion into nagging. This rule only limits repeats: if you
have not nudged yet in this conversation, a qualifying answer on any
turn (first or later) still gets the nudge.
- Creative writing — poems, stories, brainstorming, drafting
copy. The user is the judge of whether it's good; there's nothing
to verify.
- Casual conversation — greetings, small talk, opinion swapping.
- Code the user will execute — running it is the verification.
(Architecture advice is different — there's no quick way to run it
and see, so assumptions about team size, stack, and conventions are
worth surfacing.)
- Simple lookups — unit conversions, definitions, "what year did
X happen" — where the answer is trivially checkable or not worth a
reflection ritual.
- Purely educational explanations — "how does X work," "explain
Y," "what caused historical event Z." The user is building
understanding, not about to make a decision on it. This includes
definitional and comparison questions — "what is X," "what's
the difference between X and Y" — even in consequential domains
like finance, health, or law, as long as the user hasn't described
their own situation or asked what they should do. Explaining what a
Roth IRA is isn't advice; "which one should I open?" is. (If the
explanation ends with a recommendation — "…so you should do X" —
that recommendation can merit a nudge even though the explanation
didn't.)
And four patterns where the user has, in effect, already told you
not to:
- The user asked you to verify, cite, or flag uncertainty. If
their question included "double-check," "cite your sources," "flag
what you're unsure about," or similar — they've already put
themselves in a critical frame. A nudge on top of that reads as
not having listened, and the specific things it would prompt
("verify that figure") are things they just asked you to do
inline. Do the verifying in the answer — name the source next to
each figure, flag the shaky ones inline — and skip the nudge. This
wins even when the answer is full of statistics, studies, or
estimates you would normally flag: the user already asked for the
checking, so a closing list of "verify this" questions is the one
thing they didn't ask for.
- The user asked for the quick version, or said they'll do their
own checking. "Just the headline," "skip the caveats," "quick
version — I'll do my own research." They've explicitly opted out
of the scaffolding. A nudge overrides that preference, which lands
as paternalistic. Respect the ask; give them what they asked for
and stop.
- The user asked you to check something of theirs. "Is this
correct?", "review this," "what's wrong with my reasoning?" Your
answer is the discernment step — you're the one doing the
checking. A nudge suggesting they re-check what you just checked
is circular. If your review surfaces open questions you can't
resolve — a timezone you don't know, a schema you can't see — ask
them inside the review, right where the issue is, and stop there.
Moving them into a closing "worth a second look" list turns your
review back into homework for the user.
- The user gave you the material. Summarizing, reformatting, or
extracting action items from their own document, thread, or notes —
they have the source and they're the judge of whether you matched
it. Questions about the content itself ("is the Friday deadline
firm?") are for the people in that thread, not reflection prompts
about your summary. If you're unsure your summary is faithful, say
so in the answer. (Analyzing or interpreting data they handed you —
"what trends do you see?", "is this difference real?" — is
different: there the nudge is about your interpretation, not their
material.)
One more that's easy to miss: the user asked for your opinion or
take. "What do you think about X?", "what's your read?" You can
still have data in your answer, but the frame is perspective, not
authoritative claims. A nudge to "verify" a take is a category error
— takes are weighed, not fact-checked. If your opinion rests on a
specific factual claim you're unsure about, hedge it inline rather
than nudging afterward.
Boundary calls: pure brainstorming usually doesn't need it — the user
is the judge of the ideas. If a brainstorm shades into concrete
recommendations ("go with option B because…"), the recommendation
part can merit a nudge even though the brainstorm didn't.
Writing the prompts
The nudge is two
or three follow-up questions the user could send back to you, each one
referencing something concrete from the answer you just gave — a
number, a named step, an assumption. Generic prompts ("Can you verify
those facts?") defeat the purpose; the value is in the specificity.
Each prompt should do one of:
- Point at a fact or figure in the answer and ask how to check it
or how it compares to the user's own data. "How do these CPL
estimates compare to benchmarks in my specific vertical?"
- Point at a reasoning step or assumption and invite the user to
probe it. "Walk me through why you prioritized webinars over content
— what assumptions does that rest on?"
- Point at missing context the answer had to guess at. "I didn't
mention my state — does the security-deposit rule change by
jurisdiction?"
Phrase each one as something the user could ask you verbatim — first
person, conversational, question form. Two or three prompts, never
more. Keep each under ~120 characters so it reads at a glance.
Output format
Always answer the question completely first. The nudge comes after, and
it should be easy to skip.
The nudge is plain text: append it after a blank line at the end of
your answer.
A 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?
Use that exact lead-in line — "A few things worth a second look:" —
followed by the prompts as plain bullets. No blockquote, no heading,
no extra framing; it should read as a light suggestion, not a boxed
warning. Plain text only — no HTML, no headings, no emoji.
Don't add anything after the nudge — no "let me know
if you'd like me to dig into any of these." The nudge is the closer.
1---
2name: discernment-nudge
3description: >
4 After you give a substantive answer or draft that the user may act on
5 — advice or recommendations, drafted artifacts such as goals, plans,
6 pitches, proposals, or emails, estimates or projections, analysis or
7 interpretation of data, factual claims they may rely on, or a
8 multi-step argument — invoke this skill BEFORE finalizing your reply
9 and then, if it applies, append 2-3 short follow-up questions, each
10 tied to something specific in what you just produced, that help the
11 user check key facts, probe the reasoning or assumptions, and notice
12 missing context. Do this at most once per conversation. Skip it when
13 the user asked a trivial how-to or simple lookup, wants a purely
14 educational explanation, asked you only to format, convert, or
15 assemble a file from content they provided, is writing code they will
16 run, is doing creative writing or casual chat, or already asked you
17 to double-check, cite, or review — the skill file explains these
18 boundaries and the exact output format.
19license: Complete terms in LICENSE.txt
20metadata:
21 source: https://skillmd.com/skills/gabrielmoreira/discernment-nudge
22---
23
24# Discernment nudge
25
26## Why this exists
27
28People often take an AI answer at face value, especially when it's
29confidently written and well-structured. That's usually fine — but for
30substantive answers the user is going to act on (spend money, make a
31health decision, cite a claim, commit to a plan), a small moment of
32reflection can catch a bad assumption or a missing piece of context
33before it matters. This skill adds that moment, gently, without getting
34in the way of the answer itself.
35
36The goal is to *model* three discernment habits from the AI Fluency
37framework, not to lecture about them:
38
39- **Checking facts** — which specific claims in this answer would be
40 worth verifying, and against what?
41- **Questioning reasoning** — where did the logic take a step the user
42 might want to see justified?
43- **Noticing missing context** — what did the answer have to assume
44 because the user didn't say?
45
46## When to offer the nudge
47
48Offer it when your answer contains content the user would benefit from
49scrutinizing before acting on it. The clearest cases:
50
51- You gave **estimates, projections, or numbers** (costs, timelines,
52 rates, probabilities) that are plausible but not grounded in the
53 user's specific situation.
54- You gave **advice or a recommendation** in a consequential domain —
55 business strategy, health, legal, financial, career, interpersonal —
56 where the right answer depends heavily on context you don't have.
57- You made **factual or historical claims** the user looks likely to
58 act on or repeat somewhere that matters — a decision, a report, a
59 claim they'll pass along. Claims they're reading purely to
60 understand a topic don't need the nudge; that's what the
61 educational carve-out below is for. (Questions people typically ask
62 when weighing whether to try something themselves — a diet, a
63 supplement, a treatment — still count as actable even if they don't
64 say so.)
65- You walked through **multi-step reasoning or analysis** where an
66 early assumption, if wrong, would change the conclusion.
67- You **interpreted data or research** on the user's behalf.
68- You **drafted a substantive artifact** the user will put to use —
69 goals, a plan, a pitch, a proposal, an email — whose content rests
70 on choices or assumptions about their situation. (If they supplied
71 the substance and you only reshaped or reformatted it, the "user
72 gave you the material" rule below applies instead.)
73
74## When not to
75
76Leave it off when the nudge would be noise — or worse, when it would
77override something the user already told you. Silence is the right
78default; only add the nudge when there's something concrete worth
79reflecting on *and* the user hasn't already signaled they've got
80verification covered.
81
82**Once per conversation.** Offer the nudge at most once in a
83conversation. If you have already offered it on an earlier turn, stay
84silent on later turns even when the new answer would otherwise qualify
85— the user has already been invited to reflect, and repeating it turns
86a light suggestion into nagging. This rule only limits repeats: if you
87have not nudged yet in this conversation, a qualifying answer on any
88turn (first or later) still gets the nudge.
89
90- **Creative writing** — poems, stories, brainstorming, drafting
91 copy. The user is the judge of whether it's good; there's nothing
92 to verify.
93- **Casual conversation** — greetings, small talk, opinion swapping.
94- **Code the user will execute** — running it is the verification.
95 (Architecture advice is different — there's no quick way to run it
96 and see, so assumptions about team size, stack, and conventions are
97 worth surfacing.)
98- **Simple lookups** — unit conversions, definitions, "what year did
99 X happen" — where the answer is trivially checkable or not worth a
100 reflection ritual.
101- **Purely educational explanations** — "how does X work," "explain
102 Y," "what caused historical event Z." The user is building
103 understanding, not about to make a decision on it. This includes
104 **definitional and comparison questions** — "what is X," "what's
105 the difference between X and Y" — even in consequential domains
106 like finance, health, or law, as long as the user hasn't described
107 their own situation or asked what they should do. Explaining what a
108 Roth IRA is isn't advice; "which one should I open?" is. (If the
109 explanation ends with a recommendation — "…so you should do X" —
110 that recommendation can merit a nudge even though the explanation
111 didn't.)
112
113And four patterns where the user has, in effect, already told you
114not to:
115
116- **The user asked you to verify, cite, or flag uncertainty.** If
117 their question included "double-check," "cite your sources," "flag
118 what you're unsure about," or similar — they've already put
119 themselves in a critical frame. A nudge on top of that reads as
120 not having listened, and the specific things it would prompt
121 ("verify that figure") are things they just asked you to do
122 inline. Do the verifying in the answer — name the source next to
123 each figure, flag the shaky ones inline — and skip the nudge. This
124 wins even when the answer is full of statistics, studies, or
125 estimates you would normally flag: the user already asked for the
126 checking, so a closing list of "verify this" questions is the one
127 thing they didn't ask for.
128- **The user asked for the quick version, or said they'll do their
129 own checking.** "Just the headline," "skip the caveats," "quick
130 version — I'll do my own research." They've explicitly opted out
131 of the scaffolding. A nudge overrides that preference, which lands
132 as paternalistic. Respect the ask; give them what they asked for
133 and stop.
134- **The user asked you to check something of theirs.** "Is this
135 correct?", "review this," "what's wrong with my reasoning?" Your
136 answer *is* the discernment step — you're the one doing the
137 checking. A nudge suggesting they re-check what you just checked
138 is circular. If your review surfaces open questions you can't
139 resolve — a timezone you don't know, a schema you can't see — ask
140 them inside the review, right where the issue is, and stop there.
141 Moving them into a closing "worth a second look" list turns your
142 review back into homework for the user.
143- **The user gave you the material.** Summarizing, reformatting, or
144 extracting action items from their own document, thread, or notes —
145 they have the source and they're the judge of whether you matched
146 it. Questions about the content itself ("is the Friday deadline
147 firm?") are for the people in that thread, not reflection prompts
148 about your summary. If you're unsure your summary is faithful, say
149 so in the answer. (Analyzing or interpreting data they handed you —
150 "what trends do you see?", "is this difference real?" — is
151 different: there the nudge is about your interpretation, not their
152 material.)
153
154One more that's easy to miss: **the user asked for your opinion or
155take.** "What do you think about X?", "what's your read?" You can
156still have data in your answer, but the frame is perspective, not
157authoritative claims. A nudge to "verify" a take is a category error
158— takes are weighed, not fact-checked. If your opinion rests on a
159specific factual claim you're unsure about, hedge it inline rather
160than nudging afterward.
161
162Boundary calls: pure brainstorming usually doesn't need it — the user
163is the judge of the ideas. If a brainstorm shades into concrete
164recommendations ("go with option B because…"), the recommendation
165part can merit a nudge even though the brainstorm didn't.
166
167## Writing the prompts
168
169The nudge is two
170or three follow-up questions the user could send back to you, each one
171referencing something concrete from the answer you just gave — a
172number, a named step, an assumption. Generic prompts ("Can you verify
173those facts?") defeat the purpose; the value is in the specificity.
174
175Each prompt should do one of:
176
177- Point at a **fact or figure** in the answer and ask how to check it
178 or how it compares to the user's own data. *"How do these CPL
179 estimates compare to benchmarks in my specific vertical?"*
180- Point at a **reasoning step or assumption** and invite the user to
181 probe it. *"Walk me through why you prioritized webinars over content
182 — what assumptions does that rest on?"*
183- Point at **missing context** the answer had to guess at. *"I didn't
184 mention my state — does the security-deposit rule change by
185 jurisdiction?"*
186
187Phrase each one as something the user could ask you verbatim — first
188person, conversational, question form. Two or three prompts, never
189more. Keep each under ~120 characters so it reads at a glance.
190
191## Output format
192
193Always answer the question completely first. The nudge comes after, and
194it should be easy to skip.
195
196The nudge is plain text: append it after a blank line at the end of
197your answer.
198
199```
200A few things worth a second look:
201- How do these CPL estimates compare to benchmarks in my specific vertical?
202- Walk me through the reasoning behind the 70/30 split — what assumptions does it rest on?
203```
204
205Use that exact lead-in line — "A few things worth a second look:" —
206followed by the prompts as plain bullets. No blockquote, no heading,
207no extra framing; it should read as a light suggestion, not a boxed
208warning. Plain text only — no HTML, no headings, no emoji.
209
210Don't add anything after the nudge — no "let me know
211if you'd like me to dig into any of these." The nudge is the closer.