Skip to main content
skill CX Codex

communication-style — Claude Code skill: purpose and installation

The communication-style skill guides agents toward brief, scannable replies and plans. Use it when verbose explanations make output harder to review.

MIT 1 installs ★ 565 updated 2026-10-28

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add tzachbon/smart-ralph --skill communication-style -a claude-code
Open on GitHub

What communication-style does

The skill tells agents to favour brevity, even when that means using sentence fragments. It recommends bullets and tables over long prose, and asks agents to avoid repeating context or hedging. Replies follow a set sequence: a short overview, the main material, open questions, then numbered action steps.

Plans should surface unresolved questions before proposing actions. For SpecKit work, the guidance uses IDs for constitution sections and user stories, and puts file paths directly in task descriptions. The final section must be numbered steps, so the skill imposes a specific ending on agent responses.

For Claude Code, install it with npx skills add tzachbon/smart-ralph --skill communication-style -a claude-code; the skill file is listed at ~/.claude/skills/communication-style/SKILL.md. Other listed platforms include Codex, Cursor, GitHub Copilot, Gemini CLI, and OpenCode.

What it helps with

  • trim verbose agent explanations
  • shape plans into scannable lists or tables
  • place open questions before action items
  • link SpecKit tasks to file paths and story IDs

Example prompts for the agent

  • Write a concise implementation plan for authentication and finish with numbered action steps.
  • Turn this explanation into bullets, remove repeated context, and put open questions before the plan.
  • Draft a SpecKit task with its file path and user story ID.

How to install communication-style

1Skills CLI (Vercel)

npx skills add tzachbon/smart-ralph --skill communication-style -a claude-code

2Manually: put the files in place

~/.claude/skills/communication-style/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/communication-style/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about communication-style

How do I install communication-style in Claude Code?

Run npx skills add tzachbon/smart-ralph --skill communication-style -a claude-code. The skill file is at ~/.claude/skills/communication-style/SKILL.md.

What does communication-style do?

It sets rules for concise, scannable agent output. The guidance favours lists and tables, discourages repeated context, and requires numbered action steps at the end.

How does the skill guide SpecKit tasks?

It recommends referring to constitution sections and user stories by ID, with the file path included in each task description. Open questions come before the action steps.

Source text

SKILL.md · MIT
---
name: communication-style
description: Output rules for all agents - concise, scannable, actionable. Based on Matt Pocock's planning principles.
version: 0.1.0
---

# Communication Style

## Core Rule

**Be extremely concise. Sacrifice grammar for concision.**

## Why

- Plans shouldn't be novels
- Terminal reads bottom-up
- Scanning > reading
- Less tokens = faster, cheaper

## Output Rules

### 1. Brevity First

| Instead of | Write |
|------------|-------|
| "The user will be able to..." | "User can..." |
| "This component is responsible for..." | "Handles..." |
| "In order to achieve this, we need to..." | "Requires:" |
| "It should be noted that..." | (delete) |

**Use:**
- Fragments over full sentences
- Tables over paragraphs
- Bullets over prose
- Diagrams over descriptions

### 2. Structure for Scanning

Every output follows this order:

```text
1. Brief overview (2-3 sentences MAX)
2. Main content (tables, bullets, diagrams)
3. Unresolved questions (if any)
4. Numbered action steps (ALWAYS LAST)
```

### 3. End with Action Steps

**ALWAYS** end with numbered concrete steps.

```markdown
## Next Steps

1. Create auth module at src/auth/
2. Add JWT dependency
3. Implement login endpoint
4. Add tests
```

This is the LAST thing visible in terminal. Most important = most visible.

### 4. Surface Questions Early

Before action steps, list unresolved questions:

```markdown
## Unresolved Questions

- OAuth provider preference? (Google, GitHub, both)
- Session duration requirement?
- Rate limiting needed?
```

Catches ambiguities before they become bugs.

## Anti-Patterns

| Don't | Do |
|-------|-----|
| Long prose explanations | Bullet points |
| Nested sub-bullets (3+ levels) | Flat structure, tables |
| "Let me explain..." | (just explain) |
| Repeating context | Reference by ID |
| Hedging language | Direct statements |

## Examples

### Bad (verbose)

```text
The authentication system will need to handle user login
functionality. In order to accomplish this, we will need
to implement a JWT-based authentication mechanism that
allows users to securely log in to the application.
```

### Good (concise)

```text
Auth system: JWT-based login

Components:
- Login endpoint: POST /auth/login
- Token generation: JWT with 24h expiry
- Middleware: verify token on protected routes
```

## SpecKit-Specific Guidelines

### Constitution References

Use markers, not prose:
- `[C§3.1]` not "as defined in constitution section 3.1"
- `[MUST]` not "this is required by our principles"

### User Story References

Use IDs:
- `[US1]` not "the first user story about..."
- `T001 [US1]` links task to story

### Task Descriptions

Include file path inline:
- `Add auth middleware \`src/middleware/auth.ts\``
- Not "Create a new file called auth.ts in the middleware folder"

© tzachbon · MIT · tzachbon/smart-ralph

Similar skills

All skills for Agent setup: memory, context, orchestration →

Project sponsors

Companies that keep this analytics open