Agency Messaging logo

Agency Messaging

Community
chenhg5
agency-messaging

Discover all agents in the current agency and send/receive async messages between agents and the human owner via the inbox system.

Overview

Publisherchenhg5
Repositoryagencycli
Skill nameagency-messaging
Stars
240
Forks
26
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by chenhg5 on GitHub. Read the source before you install it.

Installation

Install the Agency Messaging AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/chenhg5/agencycli/tree/main/skills/agency-messaging
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/chenhg5/agencycli.git /tmp/agencycli
mkdir -p .claude/skills
cp -r /tmp/agencycli/skills/agency-messaging .claude/skills/agency-messaging
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Agency Messaging in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Agency Messaging on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Agency Messaging is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Skill: Agency Messaging

You can discover every agent in this agency and exchange async messages with them or with the human owner. Messages are non-blocking: the sender continues working immediately, and the recipient reads the message on their next wakeup.

The agency workspace is at: $AGENCY_DIR


Discover Agents

bash
agencycli --dir $AGENCY_DIR list agents
agencycli --dir $AGENCY_DIR show agent <project> <agent>

Recipient address format

  • Human owner: human
  • Any agent: <project>/<agent> — e.g. cc-connect/pm, cc-connect/qa-reviewer

Send a Message

bash
# Single recipient
agencycli --dir $AGENCY_DIR inbox send \
  --from <your-address> \
  --to   <recipient-address> \
  --subject "<subject>" \
  --body "<body>"

# Group send — repeat --to for multiple recipients
agencycli --dir $AGENCY_DIR inbox send \
  --from cc-connect/pm \
  --to cc-connect/dev-claude --to cc-connect/qa-reviewer --to human \
  --subject "Sprint kick-off" \
  --body "New sprint starts Monday. See backlog for tasks."

Examples:

bash
# PM → dev-claude: extra context
agencycli --dir $AGENCY_DIR inbox send \
  --from cc-connect/pm --to cc-connect/dev-claude \
  --subject "Issue #205 extra context" \
  --body "Only reproduces with UTF-8 filenames. Reproduce: echo '测试' > test.txt"

# PM → human: async progress update
agencycli --dir $AGENCY_DIR inbox send \
  --from cc-connect/pm --to human \
  --subject "Backlog updated" \
  --body "Added 3 new issues (P2). No action needed, just FYI."

# PM → QA: heads-up
agencycli --dir $AGENCY_DIR inbox send \
  --from cc-connect/pm --to cc-connect/qa-reviewer \
  --subject "PR incoming for #205" \
  --body "dev-claude is working on it. Expect a PR within the hour."

Reply to a Message

bash
agencycli --dir $AGENCY_DIR inbox reply <msg-id> \
  --from <your-address> \
  --body "<reply text>"

Forward a Message

bash
# Forward to a single recipient
agencycli --dir $AGENCY_DIR inbox fwd <msg-id> \
  --from <your-address> \
  --to   <recipient-address>

# Forward to multiple recipients with a note
agencycli --dir $AGENCY_DIR inbox fwd <msg-id> \
  --from cc-connect/pm \
  --to cc-connect/dev-claude --to cc-connect/qa-reviewer \
  --note "Please coordinate on this."

The forwarded message includes the original sender, subject, and body. Subject is auto-prefixed with Fwd:.


Read Messages

bash
# Your unread messages (also auto-injected into your wakeup prompt)
agencycli --dir $AGENCY_DIR inbox messages --recipient <your-address>

# Filter by sender
agencycli --dir $AGENCY_DIR inbox messages --recipient <your-address> --from human

# All messages including already-read
agencycli --dir $AGENCY_DIR inbox messages --recipient <your-address> --all

# Show archived messages
agencycli --dir $AGENCY_DIR inbox messages --recipient <your-address> --archived

# Mark all as read after listing
agencycli --dir $AGENCY_DIR inbox messages --recipient <your-address> --mark-read

Per-Message Status Management

bash
# Mark a single message as read
agencycli --dir $AGENCY_DIR inbox read <msg-id> --recipient <your-address>

# Archive (hides from normal listing, retrievable with --archived)
agencycli --dir $AGENCY_DIR inbox archive <msg-id> --recipient <your-address>

# Permanently delete
agencycli --dir $AGENCY_DIR inbox delete <msg-id> --recipient <your-address>
agencycli --dir $AGENCY_DIR inbox rm     <msg-id> --recipient <your-address>

When to Use Messaging vs. Confirm-Request

SituationUse
Need human to make a decision before you continuetask confirm-request (non-blocking, archived)
Sending info or a heads-up, no reply needed immediatelyinbox send (non-blocking)
Coordinating context between agents asynchronouslyinbox send (non-blocking)
Broadcast to multiple participants at onceinbox send --to A --to B --to C
Forwarding a message to someone elseinbox fwd
Replying to a message someone sent youinbox reply

Common PM Messaging Patterns

bash
# 1. Broadcast sprint kick-off to all agents
agencycli --dir $AGENCY_DIR inbox send \
  --from cc-connect/pm \
  --to cc-connect/dev-claude --to cc-connect/qa-reviewer --to cc-connect/biz-dev \
  --subject "Sprint W14 kick-off" \
  --body "Focus this week: <priorities>. See backlog for assigned tasks."

# 2. Notify dev of approved task
agencycli --dir $AGENCY_DIR inbox send \
  --from cc-connect/pm --to cc-connect/dev-claude \
  --subject "New task approved: <task-title>" \
  --body "Human confirmed. Priority: P<N>. Key context: <notes>"

# 3. Escalate stale task to human
agencycli --dir $AGENCY_DIR inbox send \
  --from cc-connect/pm --to human \
  --subject "Task stale: <task-title>" \
  --body "Task <id> has been in_progress for >2 days. May need intervention."

# 4. Forward a customer report from human to dev
agencycli --dir $AGENCY_DIR inbox fwd <msg-id> \
  --from cc-connect/pm --to cc-connect/dev-claude \
  --note "Customer reported this bug. Please investigate."

Frequently asked questions

What does the Agency Messaging AI skill do?

Discover all agents in the current agency and send/receive async messages between agents and the human owner via the inbox system.

Why use Agency Messaging on TypingMind?

Because you install it once and use it with any model. Agency Messaging is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Agency Messaging in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/chenhg5/agencycli/tree/main/skills/agency-messaging. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Agency Messaging?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Agency Messaging?

As many as you like. As long as a model supports skills, you can use Agency Messaging with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Agency Messaging AI skill free?

It is published on GitHub by chenhg5. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇