Discernment Nudge is an agent skill by Anthropic that adds a short reflection step after answers you might act on. It appends two or three follow-up questions tied to specific parts of the answer.
What it does
- Fires after advice, estimates, drafted plans or emails, data analysis, and factual claims you may rely on.
- Writes each question about a real figure, step, or assumption in the answer, never a generic prompt.
- Covers three habits: checking facts, questioning reasoning, and spotting missing context.
- Skips casual chat, creative writing, code you will run, simple lookups, and requests you already asked it to verify.
- Appears at most once per conversation, in a fixed plain-text format at the end.
When to use it
Use it when you want help pausing to check an AI answer before you spend money, make a decision, or pass a claim along.
How to get it
The skill is free on GitHub under the Apache 2.0 license, in Anthropic's skills repo. Anthropic says the skills in this repo are already available on paid Claude.ai plans. In Claude Code, run /plugin marketplace add anthropics/skills, then /plugin install discernment-nudge@anthropic-agent-skills.
more skills
Discernment Nudge is an agent skill by Anthropic that adds a short reflection step after answers you might act on. It appends two or three follow-up questions tied to specific parts of the answer.
What it does
- Fires after advice, estimates, drafted plans or emails, data analysis, and factual claims you may rely on.
- Writes each question about a real figure, step, or assumption in the answer, never a generic prompt.
- Covers three habits: checking facts, questioning reasoning, and spotting missing context.
- Skips casual chat, creative writing, code you will run, simple lookups, and requests you already asked it to verify.
- Appears at most once per conversation, in a fixed plain-text format at the end.
When to use it
Use it when you want help pausing to check an AI answer before you spend money, make a decision, or pass a claim along.
How to get it
The skill is free on GitHub under the Apache 2.0 license, in Anthropic's skills repo. Anthropic says the skills in this repo are already available on paid Claude.ai plans. In Claude Code, run /plugin marketplace add anthropics/skills, then /plugin install discernment-nudge@anthropic-agent-skills.
more skills
Discernment Nudge is an agent skill by Anthropic that adds a short reflection step after answers you might act on. It appends two or three follow-up questions tied to specific parts of the answer.
What it does
- Fires after advice, estimates, drafted plans or emails, data analysis, and factual claims you may rely on.
- Writes each question about a real figure, step, or assumption in the answer, never a generic prompt.
- Covers three habits: checking facts, questioning reasoning, and spotting missing context.
- Skips casual chat, creative writing, code you will run, simple lookups, and requests you already asked it to verify.
- Appears at most once per conversation, in a fixed plain-text format at the end.
When to use it
Use it when you want help pausing to check an AI answer before you spend money, make a decision, or pass a claim along.
How to get it
The skill is free on GitHub under the Apache 2.0 license, in Anthropic's skills repo. Anthropic says the skills in this repo are already available on paid Claude.ai plans. In Claude Code, run /plugin marketplace add anthropics/skills, then /plugin install discernment-nudge@anthropic-agent-skills.

