Skip to content

Honest Communication (Naval Almanack)

Helps the agent give honest, kind feedback, say no without lying, and reflect on self-deception, based on Naval Ravikant's ideas.

Stars
11,998
Licence
MIT
Files
3 files

What it does

This is a text-only guidance file that teaches an AI assistant a particular way of approaching honest conversation, drawn from a book by Naval Ravikant. When you ask how to give feedback, decline a request, or stop saying things you don't mean, the agent works from ideas such as praising specifically, criticizing ideas rather than people, and not deceiving yourself first. It then offers concrete steps like checking for one place you are fooling yourself this week. The skill's text is in Chinese, so it is most useful when you are comfortable in that language or the agent can translate for you.

Great for

  • People who want help phrasing honest feedback for a colleague, child, or friend without being hurtful
  • Anyone trying to say no to a request without making up an excuse
  • Readers who want a reflective, self-check approach to communication, in Chinese

Skip it if

  • Negotiations or situations that call for diplomatic or strategic silence, since the skill itself says full disclosure is not required and it is not meant for those cases
  • Decisions involving other people's private information or safety, because the skill says secrets belonging to others are not yours to share
  • Situations with real retaliation risk, such as hostile workplaces or authoritarian settings, where the book's advice to be fully candid may carry real costs the skill does not weigh

Try asking

  • “How do I give my coworker critical feedback on their report without hurting them?”
  • “How can I say no to a friend's favor without lying?”
  • “I think I've been saying things I don't mean in my relationship. Help me check where I'm fooling myself.”
  • “我不想说场面话,怎么真诚地给团队反馈?”
Step-by-step guide: how to use Honest Communication (Naval Almanack)Install, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 3 original files, the MIT licence and an attribution note.

    honesty-communication.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip honesty-communication.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip honesty-communication.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Describe the actual situation and who is involved so the agent can apply the 'specific praise, general criticism' idea concretely.
  • Ask for the phrasing in English or Chinese, since the source material is in Chinese.
  • Treat the output as a reflection aid, not professional counselling or legal advice on sensitive workplace or family conflicts.

What's inside

  • SKILL.md
  • test-prompts.json
  • test-results.md

Redistributed unmodified from kangarooking/cangjie-skill under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.