skill

agent-messaging — Claude Code skill: what it does, how to install

Q: How to install agent-messaging in Claude Code?

Run `npx claude-code-templates@latest --skill ai-maestro/agent-messaging` in your terminal. You must also clone `ai-maestro-plugins` and execute `./install-messaging.sh -y` to install the CLI utilities.

Enables cryptographically signed messaging between AI agents via the Agent Messaging Protocol. Useful when delegating work or managing an agent inbox.

Install into Claude Code · claude-code-templates All methods and platforms ↓

npx claude-code-templates@latest --skill ai-maestro/agent-messaging

Open on GitHub

What agent-messaging does

This skill teaches Claude Code to communicate with other agents using local AMP command-line utilities. Every dispatch receives an Ed25519 signature, keeping private keys stored on the local filesystem. Basic message exchange runs entirely on the local system without cloud dependencies.

Agent interactions rely on amp-* terminal tools. The model can create an identity with amp-init, inspect mail via amp-inbox, read payloads, or dispatch responses. Payloads can include file attachments, priority flags like urgent, and message categories such as task or handoff.

Running these operations requires the CLI utilities from the ai-maestro-plugins repository. If ~/.local/bin is missing from your PATH, Claude Code cannot call the messaging scripts.

What it helps with

send signed messages to another agent on the same machine
check and inspect incoming agent communications
route tasks and context transfers with specific priority levels
reply to existing agent message threads with file attachments

Example prompts for the agent

Check the agent inbox with amp-inbox and list unread requests
Send a task message to alice with priority urgent and attach report.pdf
Reply to message msg-42 confirming the review is finished

How to install agent-messaging

1claude-code-templates

npx claude-code-templates@latest --skill ai-maestro/agent-messaging

2Skills CLI (Vercel)

npx skills add davila7/claude-code-templates --skill agent-messaging -a claude-code

3Manually: put the files in place

~/.claude/skills/agent-messaging/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/agent-messaging/

Manually: put the files in place

.agents/skills/agent-messaging/SKILL.md

For all projects use ~/.agents/skills/

Manually: put the files in place

.cursor/skills/agent-messaging/SKILL.md

Cursor also reads .agents/skills/ and .claude/skills/

1GitHub CLI

gh skill install davila7/claude-code-templates agent-messaging

2Manually: put the files in place

.github/skills/agent-messaging/SKILL.md

Platform CLI

gemini skills install https://github.com/davila7/claude-code-templates --path cli-tool/components/skills/ai-maestro/agent-messaging

Manually: put the files in place

.devin/skills/agent-messaging/SKILL.md

The legacy .windsurf/skills/ path still works

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

Questions about agent-messaging

How to install agent-messaging in Claude Code?

Run npx claude-code-templates@latest --skill ai-maestro/agent-messaging in your terminal. You must also clone ai-maestro-plugins and execute ./install-messaging.sh -y to install the CLI utilities.

What does the agent-messaging skill do?

It provides Claude Code with commands to exchange cryptographically signed messages using the Agent Messaging Protocol. The protocol signs each transmission using local Ed25519 keys.

Does agent-messaging require an external server?

No, it operates locally on your machine by default. An external provider is only necessary when exchanging messages with external addresses like [email protected].

setup-matt-pocock-skills Matt Pocock This skill configures a repository for Matt Pocock's engineering skills by setting its issue tracker, triage labels, and domain docs layout. Run it before using the other skills. 1,052,851installs ask-matt Matt Pocock Routes you to the right Matt Pocock skill or flow for the task at hand. Use it when you need a starting point for a feature, prototype, incoming issue, or hard bug. 766,587installs to-tickets Matt Pocock Turns a plan, spec, or conversation into small end-to-end tickets with explicit blockers. Use it to agree on the work breakdown before tickets go to a tracker. 711,230installs writing-plans Jesse Vincent This skill turns requirements for a multi-step task into an implementation plan before code changes begin. It helps when another engineer needs exact files, interfaces, and checks. CX Codex 282,859installs caveman-review JuliusBrussee Turns code review comments into one line per finding: location, problem, fix. Useful when a diff or PR needs a fast pass without the filler. CX Codex +1 411,800installs caveman-compress JuliusBrussee Compresses CLAUDE.md, todo lists, and other prose files into a shorter format while retaining a readable backup. CX Codex +1 411,500installs

From the same set804

The whole set →

ai-product Guides Claude Code in shipping production LLM features, focusing on schema validation, streaming responses, token budgets, and API resilience. 250installs setup-monorepo Configures a monorepo workspace structure with build orchestration. Handy when organizing multiple packages under one repository. 250installs worktree-ghostty Splits Ghostty into three panes when a worktree is created: Claude on the left, lazygit and yazi on the right. macOS only. 249installs feature-design-assistant Turns a rough feature idea into a design and spec by asking you structured questions. Useful before a large change to the codebase. 247installs scope-guard A Claude Code hook that warns when files outside a specification scope are modified. 247installs setup-rate-limiting A slash command that inspects your project and builds a rate limiting setup: algorithms, per-user quotas, counter storage. 247installs progressive-web-app Guides Claude Code in turning web projects into installable PWAs with offline caching, web app manifests, and custom install prompts. 245installs notify-before-bash Prints a terminal notice before Claude Code executes any Bash command. Useful for keeping an eye on shell activity during coding sessions. 243installs