Discernment Nudge

Before you start: read this whole install prompt, tell me in a few lines what this skill does and what installing it involves, and wait for my go-ahead before you do anything else. Prerequisites — None. Installing this skill needs no account, no key and no payment, and using it needs nothing either: it only changes how the agent ends some of its answers. Install "Discernment Nudge" for me. After the agent gives a substantive answer I may act on (advice or a recommendation, an estimate or projection, analysis of data, a factual claim I may rely on, a multi-step argument, or a drafted plan, pitch, proposal or email), it appends two or three short follow-up questions, each tied to something specific in that answer, that I could ask back to check a fact, probe the reasoning or an assumption, or surface context the answer had to guess at. It does this at most once per conversation, under one fixed plain-text lead-in, and adds nothing after it. It stays silent for creative writing, casual chat, simple lookups and purely educational explanations, for code I will run myself, when I asked for the quick version or said I will check myself, when I already asked it to verify or cite, when I asked it to review my own work, when I supplied the material it only reshaped, and when I only asked for its opinion. Repository: https://github.com/anthropics/skills (the skill is the whole skills/discernment-nudge/ folder at the pinned commit: SKILL.md and LICENSE.txt only, 2 files) Pinned version: install the code at commit dbd4588f9e1033efb41dad4bef2f7947c8993d44 exactly. Do not substitute a newer version. If this commit is unreachable, stop and tell me. 1. Clone https://github.com/anthropics/skills into a temporary folder, check out commit dbd4588f9e1033efb41dad4bef2f7947c8993d44, and copy the skills/discernment-nudge/ folder from that checkout. Then compare every file in the folder with the fingerprints recorded for this listing on 2026-10-22: [FINGERPRINTS] - LICENSE.txt: SHA-256 bc6b3af2f331cbc7fb0da1344efb2cbe5877a31498b4d70dbc7000f3405a1362 - SKILL.md: SHA-256 9191177c4a8ef11a20dace786d708506b22d43e748c71287bb823de0dc812dad If a file does not match, it changed since the listing was prepared: do not install it yet, tell me what you found and let me decide. 2. Read SKILL.md in full, and tell me in plain words: what the skill does when used, what it connects to, whether it reads or sends any personal information or credentials, and anything that seems unusual. Stop on any red flag and tell me before going further. Confirm these points from the files: - The skill folder holds 2 files only: SKILL.md and LICENSE.txt (Apache-2.0). There is no script, no code and no configuration in it; it is a writing and judgement rule the agent follows inside the current conversation. - Nothing is sent anywhere during installation itself; the only download at install time is the skill folder from GitHub at the pinned commit. - At use time there is also nothing to send: the skill contacts no service, reads no file and no credential, and only appends two or three questions at the end of some answers, at most once per conversation. - Do not use the skill in any way during this install: using it is a separate step that only starts in later conversations, when I actually ask questions it applies to. 3. When I say "yes, install", install it as a skill: copy the whole discernment-nudge folder into the agent's skills directory, in a folder named exactly "discernment-nudge", keeping its 2 files in their original relative layout (SKILL.md at its root). 4. Verify with no network calls: frontmatter valid (name: discernment-nudge), the 2 files in place with matching fingerprints. Do not run anything from the folder and do not use the skill to verify. 5. Report what was installed, where, and what I still need to do myself (for example: nothing at all is needed from me — no account, no key, no setup; using the skill is a separate step that only starts in later conversations, and it sends nothing at any point). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.