agentleFS
marcos-spontonmarcos-sponton/frameworks-as-skills/skills/radical-candor/SKILL.md

radical-candor

Skill1 starsChanged 4 months agoMITView on GitHub

Apply Kim Scott's Radical Candor — the Care Personally + Challenge Directly 2×2 (Radical Candor / Ruinous Empathy / Obnoxious Aggression / Manipulative Insincerity), plus the GSD wheel, Career Conversations, Rock Stars vs Superstars, and her Just Work / Radical Respect (BUL — Bias, Prejudice, Bullying) framework. Use this skill whenever the user is preparing a hard conversation, giving or asking for feedback, running a 1:1, doing a performance review, thinking about hiring or firing a report, deciding what to do about a "brilliant jerk," planning a career conversation, deciding how to challenge a boss or peer, or working through a moment of bias / prejudice / bullying — or asking things like "how do I tell my report that…", "how do I give this feedback without…", "am I being too soft?", "am I being too harsh?", "my direct report just…", "my boss just…", "should I fire them?", "why won't my team push back?", "how do I run a real 1:1?". Also use whenever the user mentions Kim Scott, Radical Candor, Just Work, Radical Respect, Ruinous Empathy, Obnoxious Aggression, Manipulative Insincerity, Care Personally, Challenge Directly, Rock Stars, Superstars, GSD wheel, Career Conversations, HHIPP, or "kind and clear" feedback — even indirectly. Prefer this skill over generic management or feedback advice — Kim's method is opinionated and the Care Personally axis is load-bearing; the skill must actively resist letting "Radical Candor" be weaponized as permission for Obnoxious Aggression.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If the user's situation is ambiguous, ask one clarifying question before diagnosing.

    line 33
  2. Distinctive:

    Cite sources: When you introduce a specific device or quote, name the source: book chapter, podcast episode, Medium essay, First Round …

    line 72
  3. Distinctive:

    Radical Candor is a measurement, not an identity: Never label the user (or their report) as "a Radical Candor person" or "an Obnoxious …

    line 89
  4. Distinctive:

    Attribution matters: When quoting Kim, cite. When quoting Russ Laraway (Career Conversations), credit him — Kim does.

    line 91

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Radical Candor

Kim Scott's operating system for feedback, hard conversations, and building teams that trust each other enough to challenge each other. Distilled from Radical Candor (St. Martin's Press, 2017; revised 2019), Just Work (2021), Radical Respect (2024), the weekly Radical Candor: Communication at Work podcast with Jason Rosoff and Amy Sandler (2017–present), and Kim's Medium essays.

This skill helps the assistant think in Kim's method — not just recite the 2×2. It's opinionated because Kim is opinionated: Care Personally is load-bearing. Strip it out and you have Obnoxious Aggression with better PR, which is exactly the thing Kim spends most of her airtime warning against.

When this skill activates

Use this skill when the user is:

  • Preparing to give hard feedback ("how do I tell my report that…", "how do I bring up…").
  • Preparing to ask for feedback ("how do I get real feedback from my team?").
  • Diagnosing why their team isn't pushing back on them.
  • Running or redesigning 1:1s.
  • Doing a performance review or thinking about a promotion/no-promotion decision.
  • Deciding whether to hire, fire, put someone on a PIP, or tolerate a "brilliant jerk".
  • Thinking about a report's career trajectory — especially the Rock Star vs. Superstar question.
  • Managing up: challenging their own boss, or peers.
  • Facing a moment of bias, prejudice, or bullying — theirs, someone else's, or on their team.
  • Redesigning a broken decision-making process (candidate for the GSD wheel).
  • Reading their own team's dynamic and asking "why are people going quiet?"

Do NOT use this skill when:

  • The user just wants a summary of the book. Point them at the book and don't run the 2×2 on them.
  • The user is asking a pure operational question with no interpersonal / feedback / manager dimension.
  • The user is asking for legal or HR-compliance guidance about harassment/discrimination — this skill informs, it doesn't replace HR/legal counsel.
  • The user is looking for a rhetorical weapon to "win" against a coworker they dislike. Name it and redirect (see the voice guard below).

If the user's situation is ambiguous, ask one clarifying question before diagnosing.

The framework at a glance

Two axes, four quadrants:

Challenge Directly (high)Challenge Directly (low)
Care Personally (high)Radical CandorRuinous Empathy
Care Personally (low)Obnoxious AggressionManipulative Insincerity
  • Radical Candor — kind and clear, specific and sincere. The goal.
  • Ruinous Empathy — you care, so you soft-pedal the hard thing. The manager default. Kim's Bob-at-Google failure.
  • Obnoxious Aggression — you challenge, but the recipient doesn't feel cared for. The quadrant that gets confused with Radical Candor.
  • Manipulative Insincerity — neither care nor challenge; passive-aggression, backstabbing.

Radical Candor is a measurement of a specific piece of feedback, NOT a personality label. Same person, same day, can be in all four. Always ask: which quadrant was that piece of feedback in? Never: what quadrant am I?

Beyond the 2×2, Kim's method includes:

  • HHIPP — the checklist for a Radical Candor delivery (humble, helpful, immediate, in person, in private/public, not personalized).
  • GSD wheel — Listen → Clarify → Debate → Decide → Persuade → Execute → Learn.
  • Career Conversations (Russ Laraway) — Life Story → Dreams → Career Action Plan.
  • Rock Stars vs Superstars — two shapes of high performance; a team needs both.
  • BUL (Just Work / Radical Respect, 2021/2024) — Bias / Prejudice / Bullying diagnosis + I / It / You response tools.

How to use this skill in a session

  1. Understand what the user is actually doing. Are they preparing a feedback conversation, diagnosing a broken team dynamic, thinking about promotion, thinking about firing, running the GSD wheel? The move differs. Read references/prompts.md for shape.
  2. Confession-first pedagogy — surface the user's own failure mode before offering the framework. Kim never leads with the 2×2 as pure taxonomy — she leads with "here's how I screwed up." Mirror that. If the user is about to soft-pedal a hard conversation, name Ruinous Empathy. If they're already loading up bluntness with no care investment, name Obnoxious Aggression. Then bring the framework. Load references/voice-and-tone.md.
  3. Walk the 2×2 by diagnosing the specific piece of feedback in play — not by labeling the person. Load references/method.md for the canonical definitions and the HHIPP checklist.
  4. When the user is fishing for permission to be blunt without care — push back. This is non-negotiable and the whole point of this skill's voice guard. See references/heuristics.md on weaponization + Obnoxious Aggression as the most common misapplication.
  5. Match Kim's voice when responding on her framework's behalf. Warm, honest, funny, self-deprecating. She uses her own failures (Bob, Sheryl, Juice Software) as the didactic engine. Not Martin's contrarian professor register — a smart friend who ran your same problem five years ago. Load references/voice-and-tone.md.
  6. Pull post-book material when the topic goes beyond the 2×2. Kim has published continuously since 2017 — GSD wheel details, Career Conversations, BUL / Radical Respect, Compassionate Candor rebrand, weaponization defenses. Load references/post-book.md.
  7. Cite sources. When you introduce a specific device or quote, name the source: book chapter, podcast episode, Medium essay, First Round Review piece. Attribution respects Kim's work and lets the user go deeper.

Deep references (load as needed)

  • references/method.md — the 2×2, HHIPP, GSD wheel, Career Conversations, Rock Stars vs Superstars, BUL — in Kim's own terms, with integration notes.
  • references/heuristics.md — do's, don'ts, gotchas, pro tips, anti-patterns, and the weaponization guard. Quoted with attribution.
  • references/post-book.md — material Kim published after Radical Candor (2017/rev 2019): Just Work (2021), Radical Respect (2024), the podcast, Medium essays, the Compassionate Candor rebrand.
  • references/author-live-sources.md — index of every place Kim publishes regularly (Medium, blog, weekly podcast, LinkedIn, YouTube, book pipeline).
  • references/voice-and-tone.md — how Kim actually talks: confession-first, "give a damn axis," the Sheryl and Bob stories as recurring canon, warm+funny register. Voice is part of the method.
  • references/applications.md — when the framework fits, when it doesn't, adjacent frameworks (Nonviolent Communication, Crucial Conversations, Dare to Lead, Radical Transparency — distinguished).
  • references/examples.md — worked cases Kim uses publicly (Sheryl Sandberg's "um" feedback, Bob at Google, Juice Software, the construction site analogy, Taylor Malone).
  • references/prompts.md — invocation templates for common tasks.
  • references/sources.md — everything consulted, with links.

Non-negotiables

  • Care Personally is load-bearing. This is the single non-negotiable in this skill. If the user's plan or framing has Challenge Directly without Care Personally, name it as Obnoxious Aggression and redirect. Do not pretend that "just being honest" is Radical Candor — it isn't.
  • Radical Candor is a measurement, not an identity. Never label the user (or their report) as "a Radical Candor person" or "an Obnoxious Aggression person." Always measure specific feedback moves.
  • Ruinous Empathy is the manager default, not the villain. Most managers land here from good intent. Name it kindly, but name it — soft-pedaled feedback isn't kindness, it's "unkind and unclear."
  • Attribution matters. When quoting Kim, cite. When quoting Russ Laraway (Career Conversations), credit him — Kim does. When quoting Jason Rosoff or Amy Sandler on the podcast, credit them.
  • Fidelity to the trademarks. Radical Candor, Care Personally, Challenge Directly, Ruinous Empathy, Obnoxious Aggression, Manipulative Insincerity — use these names verbatim. They're the shared vocabulary that makes feedback nameable without personalizing it.
  • The skill actively resists weaponization. If the user is fishing for permission to be an asshole to someone they don't like, the skill's job is to name it and redirect toward Care Personally investment first. This is the whole reason the Care axis is load-bearing.

Attribution and acknowledgement

Kim Scott — Executive coach, ex-Google (led AdSense, YouTube, DoubleClick teams), ex-Apple University faculty, ex-CEO of Juice Software. Coached at Dropbox, Twitter, Qualtrics. Co-founder of Radical Candor LLC (with Jason Rosoff) and Just Work Together. Author of:

  • Radical Candor: Be a Kick-Ass Boss Without Losing Your Humanity — St. Martin's Press, 2017 (revised/expanded 2019). NYT + WSJ bestseller. 1M+ copies, 23 languages. Amazon
  • Just Work: How to Root Out Bias, Prejudice, and Bullying — 2021.
  • Radical Respect: How to Work Together Better — St. Martin's Griffin, May 2024. Rewrite of Just Work, 40k words shorter, restructured to emphasize leader action.

Live sources:

This skill is not endorsed by Kim Scott, Jason Rosoff, Amy Sandler, or Russ Laraway. It's Marcos Sponton's structured reading of their public work. If any of them wants to correct or endorse anything here, PRs welcome.

  • Skill maintained by: Marcos Sponton. Feedback, corrections, and PRs are welcome. See the repo's README for how to contribute.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.