Agent skill

Internal Comms

by borghei in borghei/Claude-Skills

Plan, sequence, and pressure-test internal announcements before they send.

MITAuto-check passedWriting & Content

Install Internal Comms

skills CLI
$ npx skills add borghei/Claude-Skills --skill internal-comms -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install borghei/Claude-Skills internal-comms --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/borghei/Claude-Skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/business-operations/internal-comms .claude/skills/internal-comms && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
internal-comms
GitHub stars
874
Token cost
~2.9k tokens
SKILL.md length
1,477 words
Files
11 (incl. scripts, references, assets)
Skills in repo
364
Repo updated
First seen
Licence
MIT

At a glance

Plan, sequence, and pressure-test internal announcements before they send.

  • Works in 4 steps: Save the draft as JSON with title, body,… → Run the auditor. It scores five required… → Fix every FAIL element. Treat WARN on… → …
  • Announcing a reorg
  • SKILL.md covers When to use this skill, Inputs the skill expects, Clarify First and Workflows, plus 3 more sections
  • Runs Python scripts from its folder; calls python3

What it does

Internal Comms is an agent skill from borghei/Claude-Skills. Plan, sequence, and pressure-test internal announcements before they send. Use when announcing a reorg, policy, or product change, drafting an all-hands or exec update, or deciding who hears what and in which order.

Its SKILL.md is about 2.9k tokens, which your agent loads only when the skill is triggered. The skill folder holds 13 other files, including scripts, reference files and assets (for example `assets/announcement-template.md`, `assets/comms-plan-template.md` and `assets/sample_announcement.json`).

It sits in Writing & Content, covering Internal communications. The repository describes itself as: 385 AI skills, 77 expert agents, and 900 stdlib Python tools for every team: engineering, PM, marketing, C-level, compliance, business ops, research, and a LinkedIn toolkit… The licence is MIT.

When your agent uses it

  • Announcing a reorg
  • Drafting an all-hands
  • Deciding who hears what and in which order

Example prompts

  • “/internal-comms”

Requirements

  • Python 3

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

  1. Save the draft as JSON with title, body, audience, and channel.
  2. Run the auditor. It scores five required elements (what changed, why, who is affected,
  3. Fix every FAIL element. Treat WARN on reading grade as advisory unless the audience
  4. Re-run until the score clears 75.

What it can do on your machine

Read from SKILL.md and the folder at commit c9a1487. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 3 files in scripts/ (Python), which the agent can run.

    Shell commands in SKILL.md call:

    • python3

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Internal Comms loads about 2.9k tokens when it runs, and up to ~7.7k if it reads all its reference files. Until then it costs about 58 tokens; SKILL.md has 1,477 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~58
When it runs · the whole SKILL.md, loaded when a task matches
~2.9k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~7.7k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from borghei/Claude-Skills at commit c9a1487, republished under its MIT licence (© borghei). 1,477 words, ~2,938 tokens.

Download SKILL.mdSave it as .claude/skills/internal-comms/SKILL.md (or your agent's skills folder). This skill also uses 10 other files; get the full folder from GitHub.
name
internal-comms
description
Plan, sequence, and pressure-test internal announcements before they send. Use when announcing a reorg, policy, or product change, drafting an all-hands or exec update, or deciding who hears what and in which order.
license
MIT + Commons Clause
metadata.version
1.0.0
metadata.author
borghei
metadata.category
business-operations
metadata.domain
internal-communications
metadata.updated
2026-07-21
metadata.tags
internal-comms, change-management, announcements, all-hands, stakeholder-sequencing

Internal Communications

Internal communication fails in predictable ways: the wrong people hear it first, the message explains the decision but not the consequence, and nobody tests the draft before it reaches 2,000 inboxes. This skill treats an announcement as an artifact with a blast radius, a sequence, and a pass/fail quality bar — not as a writing exercise.

The three levers are who hears it in what order, what the message must contain, and which channel carries it. Get those right and tone matters far less than people assume.

When to use this skill

  • Announcing a reorg, layoff, or leadership change where sequencing errors are unrecoverable
  • Rolling out a policy change (RTO, expenses, security, performance process) that people must act on
  • Communicating a product or system migration with a deadline and a required user action
  • Structuring a recurring all-hands or weekly exec update that currently reads as a status dump
  • Pressure-testing a draft before send: reading level, jargon density, missing elements
  • Deciding channel and cadence for a multi-week change programme rather than a one-shot email

Inputs the skill expects

  • The change itself: what is changing, effective date, whether it is reversible
  • Audience list with impact level (high / medium / low) and rough headcount per group
  • Draft text, if one exists (plain markdown or text is fine)
  • Constraints: embargo, legal review, regulatory disclosure, market-sensitivity
  • Available channels and which ones your org actually reads
  • Who can answer questions after send, and where

Clarify First

Before generating, confirm these inputs. If any is unknown or vague, ASK — do not assume:

  • Blast radius — who is materially affected vs merely informed — this decides whether you need a cascade (manager-first) or a single broadcast, and cascades cost 3-5 days
  • Reversibility of the decision — reversible changes can be announced as proposals with a feedback window; irreversible ones must never be, because inviting input you will ignore is the fastest way to lose trust
  • Effective date and the required user action — sets every T-offset in the plan and determines whether the message is informational or instructional
  • Whether anything is legally or market-sensitive — embargo constraints override the ideal sequence and force simultaneous rather than cascaded delivery

Stop rule: ask only the 2-3 that most change the output. If the user says "just draft it," proceed and list your assumptions at the top of the artifact.

Workflows

Workflow 1 — Audit a draft before it sends

Run this on every announcement that reaches more than ~50 people. It takes 30 seconds and catches the failure modes that generate the most follow-up questions.

  1. Save the draft as JSON with title, body, audience, and channel.
  2. Run the auditor. It scores five required elements (what changed, why, who is affected, what to do, where to ask), reading grade, jargon density, and hedging.
  3. Fix every FAIL element. Treat WARN on reading grade as advisory unless the audience includes non-native speakers or frontline staff, in which case treat it as FAIL.
  4. Re-run until the score clears 75.
bash
python3 business-operations/internal-comms/scripts/announcement_auditor.py \
  --input business-operations/internal-comms/assets/sample_announcement.json \
  --min-score 75 --format text
Workflow 2 — Sequence a change communication
  1. Describe the change and list audiences with impact and headcount.
  2. Run the sequencer. It computes blast radius, orders audiences by a impact-then-influence rule, and assigns T-offsets in days relative to broadcast.
  3. Review the manager-enablement step — if the plan says you need one and you skip it, managers learn about the change from their own reports, which is the single most common reorg comms failure.
  4. Export the plan and assign an owner per step.
bash
python3 business-operations/internal-comms/scripts/comms_sequencer.py \
  --input business-operations/internal-comms/assets/sample_change.json \
  --format text
Workflow 3 — Pick the channel
  1. Characterise the message on six dimensions: urgency, complexity, audience size, sensitivity, whether it needs two-way discussion, and whether it needs a durable record.
  2. Run the scorer to rank channels and see which are actively disqualified.
  3. Use the top-ranked channel as the primary, and the highest-ranked durable channel as the system of record. Never let a chat message be the record.
bash
python3 business-operations/internal-comms/scripts/channel_fit_scorer.py \
  --input business-operations/internal-comms/assets/sample_message_profile.json \
  --top 4 --format json

Decision frameworks

Blast radius → sequence pattern [PROVEN]

Blast radius is the count of people whose work or employment changes, not the count who receive the email.

RadiusPatternLead time before broadcastManager enablement
1-10Direct 1:1 conversations, then quiet team noteSame dayNot needed
11-50Manager brief → team meetings → written follow-up1-2 daysRequired, 60-min session
51-250Exec brief → manager brief + FAQ → all-hands → written3-5 daysRequired, with talk track + FAQ
250+Add a pre-brief for informal influencers and a 2-week reinforcement cadence5-10 daysRequired, plus a manager Q&A channel

Escape hatch: legal embargo or market sensitivity collapses this to a simultaneous broadcast. When that happens, say so in the message — "we could not brief managers first because of disclosure rules" — rather than letting people infer carelessness.

The five required elements [PROVEN]

Every announcement of consequence answers these, in this order. A message missing any one of them generates a predictable class of follow-up question.

ElementMissing it producesMinimum bar
What changed"Wait, is this happening now or later?"One sentence, in the first paragraph, with the effective date
WhyRumour and worst-case theoryThe actual reason, including the constraint that forced it
Who is affectedEvery reader assumes it is themNamed groups, plus an explicit "if you are not in these groups, nothing changes for you"
What to doNothing happensAn action with a deadline, or the words "no action needed"
Where to askQuestions go to the wrong peopleA named human or channel, plus when they will respond
Show full SKILL.md (576 more words)Show less
TypeLead withLengthFeedback window
Irreversible decision (reorg, layoff, shutdown)The decision300-500 wordsNone — do not solicit input
Reversible policyThe problem being solved400-700 words5-10 business days
Migration / deadlineThe action and the date200-350 wordsQuestions only
Celebratory / milestoneThe outcome and who did it150-250 wordsNone
Bad news (incident, miss)What happened, in plain words250-400 wordsPost-mortem link

Announce once and people forget; announce weekly and they tune out. The pattern that holds: T-0 announcement, T+3 days FAQ update, T+1 week manager check-in, T+2 weeks progress note, T+4 weeks close-out. Five touches, decreasing in length. See references/change-comms-cadence.md for the full grid including reinforcement channels.

Anti-Patterns

The Broadcast Ambush

Mistake: Sending a reorg or policy announcement to all-staff without briefing managers first. Why it happens: Leadership fears leaks, and a simultaneous send feels "fair" and egalitarian. Instead: Brief managers 24-48 hours ahead with a talk track and an FAQ, and tell them explicitly what they may and may not repeat. A manager who cannot answer their team's first question loses standing that takes months to rebuild. If leak risk is genuinely severe, compress the window to 2 hours rather than eliminating it.

Burying the Decision Under the Reasoning

Mistake: Opening with three paragraphs of market context before stating what is changing. Why it happens: The author wants the reader to reach the same conclusion they did, so they reconstruct their own reasoning path. Instead: State the decision in sentence one, then the reasoning. Readers who disagree will read the reasoning more carefully, not less. Context-first structure reads as though you are building a case, which signals defensiveness.

The Fake Consultation

Mistake: Announcing a settled decision with "we'd love your feedback" attached. Why it happens: It softens the delivery and feels more collaborative than a flat directive. Instead: If the decision is settled, say so: "This is decided. Here is what we considered." If input can genuinely change the outcome, say what specifically is still open and by when. Solicited-then-ignored feedback is measurably worse for trust than no consultation at all.

Jargon as Cushioning

Mistake: "We are realigning our operating model to better leverage synergies across pods." Why it happens: Abstraction makes uncomfortable news feel less blunt to the person writing it. Instead: Write the sentence a person would say out loud. The auditor flags jargon density above 2% of words; anything above that on a high-impact message means the draft is hiding something, and readers will assume the worst thing it could be hiding.

No Named Owner for Questions

Mistake: Ending with "reach out with any questions." Why it happens: Nobody wants to volunteer their calendar. Instead: Name a person or a specific channel and a response commitment ("#change-rto, answered within one business day"). Unrouted questions become hallway conversations, and hallway conversations become the version of the message people remember.

Files

FilePurpose
scripts/announcement_auditor.pyScores a draft on required elements, reading grade, jargon, hedging, and structure
scripts/comms_sequencer.pyBuilds an ordered, T-offset communication plan from a change description and audience list
scripts/channel_fit_scorer.pyRanks communication channels against a six-dimension message profile
references/announcement-archetypes.mdEight announcement archetypes with structure, length, and worked openings
references/change-comms-cadence.mdTiming grids, reinforcement schedules, and manager-enablement content
assets/announcement-template.mdFill-in template enforcing the five required elements
assets/comms-plan-template.mdSequencing plan table with owner, channel, and T-offset columns
assets/sample_announcement.jsonRunnable input for announcement_auditor.py
assets/sample_change.jsonRunnable input for comms_sequencer.py
assets/sample_message_profile.jsonRunnable input for channel_fit_scorer.py

© borghei, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 10 other files (scripts, references, assets) in business-operations/internal-comms of borghei/Claude-Skills.

  • SKILL.md
  • assets/announcement-template.md
  • assets/comms-plan-template.md
  • assets/sample_announcement.json
  • assets/sample_change.json
  • assets/sample_message_profile.json
  • references/announcement-archetypes.md
  • references/change-comms-cadence.md
  • scripts/announcement_auditor.py
  • scripts/channel_fit_scorer.py
  • scripts/comms_sequencer.py

Open the folder on GitHubat commit c9a1487

Compare with similar skills

Internal Comms next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Internal Comms compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Internal Comms this skillborghei/Claude-Skills874—~2.9kAutomated safety check: PassMIT
Internal Communications Writeranthropics/skills180k37 repos~378Automated safety check: PassApache-2.0
Natural Japanese Business Writingcoji/natural-japanese1.9k—~2.1kAutomated safety check: PassMIT
My Weekly Reportmeain/dotfiles285—~2.5kAutomated safety check: PassMIT
Updatesjtenniswood/espframe209—~527Automated safety check: PassCustom licence
Updateakseolabs-seo/AK-Threads-booster275—~1.1kAutomated safety check: NotesMIT

Similar skills

  • Official

    A set of resources to help me write all kinds of internal communications, using the formats that my company likes to use. Claude should use this skill…

    180k GitHub starsUsed in 37 repos~378 tokens
    Writing & ContentAuto-check passed
  • Writes and edits Japanese business documents so they read clearly and naturally, removes AI-sounding phrasing and can score how AI-like a text reads.

    1.9k GitHub stars~2.1k tokensUpdated 1 mo ago
    Writing & ContentAuto-check passed
  • My Weekly Report

    meain/dotfiles

    Generate a concise weekly status update in team format. An agent skill from meain/dotfiles.

    285 GitHub stars~2.5k tokensUpdated 1 mo ago
    Writing & ContentAuto-check passed
  • Updates

    jtenniswood/espframe

    Summarize user-facing changes made to this project over the last 7 days.

    209 GitHub stars~527 tokensUpdated yesterday
    Writing & ContentAuto-check passed
  • Update

    akseolabs-seo/AK-Threads-booster

    Check AK-Threads-Booster for upstream GitHub updates, safely fast-forward the local skill repo, or install an opt-in weekly Codex automation that keeps the skill on the latest version.

    275 GitHub stars~1.1k tokensUpdated 3 mo ago
    Writing & ContentAuto-check: notes
  • Weekly Status Report

    sgharlow/claude-code-recipes

    Turn scattered weekly updates (notes, emails, task exports, meeting notes) into a structured status report with highlights, progress, severity-rated blockers, and next-week priorities.

    388 GitHub stars~602 tokensUpdated 2 mo ago
    Writing & ContentAuto-check passed

More from borghei/Claude-Skills

All 364 skills in this repo
  • Agents In The Team

    borghei/Claude-Skills

    Run delivery when AI coding and ops agents take tickets. An agent skill from borghei/Claude-Skills.

    874 GitHub stars~4.2k tokensUpdated yesterday
    Auto-check passed
  • AI Content Disclosure

    borghei/Claude-Skills

    Check AI-generated marketing content and reviews for required disclosures under the EU AI Act, FTC rules and platform AI-label policies.

    874 GitHub stars~3.4k tokensUpdated yesterday
    Auto-check passed
  • AI Prototyping

    borghei/Claude-Skills

    Idea to AI-generated prototype to customer validation to engineering handoff.

    874 GitHub stars~3.6k tokensUpdated yesterday
    Auto-check passed
  • Analytics Engineer

    borghei/Claude-Skills

    Analytics engineering across data modeling, dbt, transformation, and semantic layers.

    874 GitHub stars~3.4k tokensUpdated yesterday
    Auto-check passed
  • Ansoff Matrix

    borghei/Claude-Skills

    Ansoff Matrix — 4-quadrant framework for growth options: market penetration, market/product development, and diversification.

    874 GitHub stars~2.2k tokensUpdated yesterday
    Auto-check passed
  • Brainstorm Okrs

    borghei/Claude-Skills

    OKR brainstorming and validation using the Radical Focus framework — outcome objectives, measurable key results, counter-metrics.

    874 GitHub stars~1.4k tokensUpdated yesterday
    Auto-check passed

Questions about Internal Comms

What does Internal Comms do?

Plan, sequence, and pressure-test internal announcements before they send. Internal Comms is an agent skill from borghei/Claude-Skills. Plan, sequence, and pressure-test internal announcements before they send.

When should I use Internal Comms?

Internal Comms fits situations like: announcing a reorg; drafting an all-hands; deciding who hears what and in which order.

How do I install Internal Comms in Claude Code?

Run `npx skills add borghei/Claude-Skills --skill internal-comms -a claude-code`. Or copy the skill folder (business-operations/internal-comms in borghei/Claude-Skills) into .claude/skills/internal-comms in your project. Claude Code loads it when a task matches its description.

How do I install Internal Comms in Codex?

Run `npx skills add borghei/Claude-Skills --skill internal-comms -a codex`. Or copy the skill folder (business-operations/internal-comms in borghei/Claude-Skills) into .agents/skills/internal-comms in your project. Codex loads it when a task matches its description.

Can I use Internal Comms in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add borghei/Claude-Skills --skill internal-comms -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/internal-comms, .gemini/skills/internal-comms, .github/skills/internal-comms and .opencode/skills/internal-comms in your project.

What does Internal Comms need to run?

Going by SKILL.md and its folder, Internal Comms needs Python for the scripts in its folder and the command-line tools its instructions call (python3). Our summary lists: Python 3.

Does Internal Comms access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Internal Comms safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Internal Comms use?

Internal Comms is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Internal Comms use?

About 2.9k tokens (SKILL.md is roughly 12k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 4.8k tokens, read only when the agent opens those files.

What are the alternatives to Internal Comms?

Skills that share tags, products or a category with Internal Comms: Internal Communications Writer (anthropics/skills, 180k stars), Natural Japanese Business Writing (coji/natural-japanese, 1.9k stars), My Weekly Report (meain/dotfiles, 285 stars) and Updates (jtenniswood/espframe, 209 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Internal Comms?

borghei (a GitHub user) maintains it in borghei/Claude-Skills, which has 874 GitHub stars. The repository holds 364 skills in this directory. The repository was last updated on October 7, 2026.

Source: borghei/Claude-Skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.