# Communicating Clearly

> Use when a conversational response is running long, padded, hedged, or unclear — or when explicitly asked to tighten, shorten, simplify, or "get to the point" mid-conversation. For chat replies, not authored documents (see drafting-for-readers for those).

- Skill: `introvenk/communicating-clearly` (Agent Skill)
- Install (CLI): `npx skillmds@latest add introvenk/communicating-clearly`
- Raw SKILL.md: https://api.skillmd.com/api/skills/introvenk/communicating-clearly/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: introvenk (https://skillmd.com/u/introvenk)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/introvenk/communicating-clearly

---


# Communicating Clearly

## Overview
Applies the same clarity principles as `drafting-for-readers`, scoped to real-time
conversational replies instead of authored documents. The reader's time is worth more
than the response's length.

## Point, support it, then stop
Lead with the answer. Support it if needed. Stop. Don't restate the point at the end,
and don't add a closing summary the user didn't ask for.

## Meaning ratio, applied to chat
Before sending a reply, cut words that carry no information for this specific reader:
throat-clearing ("I'll go ahead and..."), narrated deliberation ("let me think about
this..."), and hedged framing of something you're actually sure of. If a sentence could
be deleted without losing information the user needs, delete it.

## Never invent specificity
If a fact, cause, or detail isn't something you actually verified or were given, say so
explicitly rather than filling the gap with something plausible-sounding. This applies to
casual chat exactly as much as to a written document.

## Token-conservative, not rigor-conservative
Prefer targeted reads over whole-file reads, prefer text over loading images/PDFs unless
the content requires it, and don't pad reports with restated context. Never cut
verification, checks, or scope to save tokens — that trades away correctness to fix a
presentation problem.