Psychological Safety

---
name: psychological-safety
description: Use when a team lacks psychological safety — members don't speak up, withhold concerns, or fear making mistakes. Triggers include "build psychological safety", "team is afraid to speak up", "create safe environment", "people are withholding their opinions", "no one challenges decisions", "fear-based culture". Prescriptive practices from Amy Edmondson's research — not just diagnostic like `team-diagnosis`, but what to do this week.
---

# Psychological Safety Builder

Build psychological safety — the belief that one can speak up with ideas, questions, concerns, or mistakes without punishment or humiliation. From Amy Edmondson (via Google's Project Aristotle): the #1 factor in team effectiveness. **Not** "everyone is nice" (healthy conflict is safe), "no accountability" (high standards + safety = high performance), or "no fear" (safety means fear doesn't silence you). **Is:** interpersonal risk-taking is rewarded, not punished.

## When to apply

Trigger when: members stay silent in meetings then complain afterward; no one admits mistakes; no one challenges the leader; retros are superficial; or the user says *"I need my team to tell me what's really going on"*. Skip if the team already demonstrates safety or the user wants a diagnostic only (use `team-diagnosis`).

## The four pillars

1. **Leader vulnerability** 2. **Response to bad news** 3. **Inclusion and belonging** 4. **Enabling conditions** — rituals that reinforce safety day-to-day.

## The leader's actions (this week)

### 1. Model vulnerability (Pillar 1)
- Say: *"I was wrong about X. Here's what I missed."*
- Ask questions you don't know the answer to; say "I don't know" instead of guessing.
- Share a recent mistake and what you learned.
- Solicit feedback: *"What should I start, stop, or continue?"* — and listen.

**Frequency:** 2–3 times per week until habitual.

### 2. Respond constructively to upward speech (Pillar 2)
When someone raises a concern or admits a mistake: **thank them immediately**; **listen without interrupting**; separate message from delivery; **take action visibly**. **Never:** shoot the messenger, minimize, get defensive, ignore, or punish. **Practice:** role-play the default reaction and replace it.

### 3. Actively solicit and honor all voices (Pillar 3)
- **Round-robin:** "Let's hear from everyone before I summarize" — call on the quiet ones by name.
- **Pre-meeting or anonymous input** for sensitive topics.
- **Disrupt interruptions:** "Let Alice finish."
- **Value dissent** and **credit publicly** ("Carol was right").

**Measure:** track participation distribution; aim for balanced airtime.

### 4. Create enabling structures (Pillar 4)
- **Pre-mortems:** everyone writes 1–2 risks anonymously, then discusses.
- **Blameless post-mortems:** "what went wrong", not "who messed up" — use `the-postmortem-facilitator` agent if available.
- **Also:** appreciation rounds, decision debriefs, no-interruption policy, decisions documented with rationale, quarterly assumption testing, safe escalation path.

## Assessing current safety level

Quick diagnostic (not a substitute for `team-diagnosis`) — **anonymous 5-question pulse (1–5 scale):**
1. If I make a mistake on this team, it is used against me
2. It is safe to speak up and disagree with my teammates
3. Problems and concerns can be raised without fear
4. My unique skills and perspectives are valued
5. This team is inclusive; I feel I belong

Average ≤ 3 indicates significant safety work needed. **Signals:** silence after controversial points; "meetings after the meeting"; no admitted errors in retros.

## Common blockers

- **"Our leader is the problem":** leader runs this skill on themselves, solicits anonymous 360 feedback, models vulnerability publicly; escalate via `stakeholder-reflect` on the leader if no change.
- **"Too busy for rituals":** safety is risk reduction — start with a 5-minute pre-mortem.
- **"We already trust each other":** *"When did someone last openly disagree with the leader?"* "Never" means safety is lower than perceived.
- **"Conflict is toxic here":** encourage **task conflict**, suppress **relationship conflict** — "Attack ideas, not people."

## Integrating with the 3X3 team framework

- **Team Charter:** explicit safety norm: *"We assume good intent. We speak up early. Mistakes are learning opportunities."*
- **Team Check-in:** pulse: *"How safe did you feel speaking up this sprint (1–5)?"*
- **Team Close-gap:** small step: *"Each of us shares one mistake in retro without judgment."*

## Leader self-check (weekly)

Did I model vulnerability? Reward or punish the last person who spoke up? Give all voices airtime? Surface bad news early? Two or more "no" → repair conversation: *"I noticed X. Here's what I'll do differently."*

## Operating principles

- **Safety ≠ comfort.** You can be safe and challenged.
- **It's the leader's job to build it** — safety is designed, not accidental.
- **Start small** — one admission, one thanked concern, one pre-mortem compound.
- **Repair quickly** — within 24h.

## Composition with other skills

- **`team-check-in`** — includes safety pulse; if low, run this skill
- **`team-style-inventory`** — style mismatches undermine safety
- **`read-the-room`** — detect hidden safety issues in meetings
- **`coaching-mode`** — develop the leader's own vulnerability
- **`feedback-frameworks`** — teach the team to give feedback safely

## Anti-patterns to flag

- **Safety theater:** "We're all friends here" while dissent is punished.
- **Safety as permissive:** no standards — that's complacency.
- **One-off events:** one offsite doesn't build lasting safety.
- **Ignoring exclusion:** one person consistently interrupted is a safety problem.
- **Surveillance:** track to ensure voices are *invited*, not to catch silent people.

## Further reading

Amy Edmondson — "The Fearless Organization"; Google's Project Aristotle; Lencioni's "Five Dysfunctions".