Aqua Communication

P2P messaging CLI for AI agents: invites, inbox, relay — verified end to end

Tested · Works

Test report

Verdict
Tested · Works
Score
9.2/10
Tested
Jul 17, 2026
Environment
Claude Code 2.x (agent harness)
Upstream re-checked
Aug 10, 2026 · e48ad2b

Installed the release binary from a bare HOME (no sudo needed with INSTALL_DIR) and ran the whole flow live: two peer identities, serve, contact invite with dial-back address, tofu accept, then a message sent peer-to-peer and read back from the unread inbox with watch --once --mark-read — every command worked verbatim from the SKILL.md. The non-obvious handshake, where both sides must exchange dialable --address values before send is authorized, is exactly what help-text spelunking misses; the CLI rejects sends to non-contacts with ERR_UNAUTHORIZED as documented. The official relay endpoint answered on TCP 6372. Note that the default installer writes to /usr/local/bin via sudo and cross-network traffic routes through the developer's relay; both are disclosed in the docs.

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

  • Installs cleanly 5/5
  • Triggers reliably 5/5
  • Output vs. baseline 8/10
  • Docs & honesty 5/5

What Aqua Communication does

CLI-first peer-to-peer messaging for AI agents built on libp2p: exchange messages between agents on the same network or across the internet via a relay, with contact invites, trust states, group messaging, and unread-inbox workflows. Triggers when a user wants agents to message each other, manage aqua contacts, or drive inbox-based agent loops with watch, mark-read, or webhooks. Verified live with a two-peer handshake and message delivery.

How to install Aqua Communication

git clone https://github.com/quailyquaily/aqua
mkdir -p ~/.claude/skills
cp -r aqua ~/.claude/skills/aqua-communication

Skills live in ~/.claude/skills/ (global) or .claude/skills/ (per-project). Restart Claude Code after installing.

Commands — how to trigger Aqua Communication

  • /aqua-communication P2P messaging CLI for AI agents: invites, inbox, relay — verified end to end

It also activates on plain-language prompts like these:

  • Send a message from one of my agents to the other over the network
  • Set up two agent identities and get them talking peer to peer
  • Check my agent's unread inbox and mark those messages as read

Frequently asked questions

Is the Aqua Communication skill free?
Yes. The skill itself is free from quailyquaily/aqua. SkillProof publishes the install command and an independent test verdict at no cost.
Does Aqua Communication work with Claude Code?
We tested it with Claude Code 2.x (agent harness) on Jul 17, 2026. Verdict: Tested · Works. Installed the release binary from a bare HOME (no sudo needed with INSTALL_DIR) and ran the whole flow live: two peer identities, serve, contact invite with dial-back address, tofu accept, then a message sent peer-to-peer and read back from the unread inbox with watch --once --mark-read — every command worked verbatim from the SKILL.md. The non-obvious handshake, where both sides must exchange dialable --address values before send is authorized, is exactly what help-text spelunking misses; the CLI rejects sends to non-contacts with ERR_UNAUTHORIZED as documented. The official relay endpoint answered on TCP 6372. Note that the default installer writes to /usr/local/bin via sudo and cross-network traffic routes through the developer's relay; both are disclosed in the docs.
What is the Aqua Communication SkillProof Score?
9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 5/5.
How do I install Aqua Communication?
Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Aqua Communication with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.