1panel-dev/1claw/ct-chat

Claude Code and agent skill by 1panel-dev/1claw

Trust tier 2
  • community
LLM review pending
Source check pending

Checking that the SKILL.md file is still in the repository.

Use this skill when an Agent needs to send a tracked CT message to a concrete CT ID. Do not use it for ordinary chat or when the target is still unknown.

Install

npx vskill@latest install 1panel-dev/1claw/ct-chat

SKILL.md

CT Chat

In this skill, the following names refer to the same sending action:

  • Claw Team Channel
  • claw-team
  • CT Channel
  • CT
  • CT Call

When To Use

Use this skill when all of the following are true:

  1. The current Agent needs collaboration, confirmation, execution, review, notification, or acceptance follow-up through a tracked CT message.
  2. The collaboration target can be identified by a concrete target CT ID.
  3. The request has a clear topic and a concrete ask.

Do not use this skill when:

  • the current Agent can finish the work alone
  • the target is still unclear
  • the request is vague and has not been narrowed into a concrete ask
  • an ordinary reply in the current conversation is enough

How To Use

Follow this workflow:

  1. Confirm that a tracked CT message is really needed.
  2. Identify the target CT ID.
  3. Narrow the request into one clear topic and one clear message.
  4. Use the claw-team channel to send the structured payload.
  5. Let Claw Team track the result so the human can observe or intervene.

Target Rules

Preferred target format:

  • CTA-0009
  • CTU-0001

Also accepted by the channel:

  • ctid:CTA-0009
  • @CTA-0009

Use the plain CT ID form unless you have a strong reason not to.

Payload Contract

Current first-stage payload:

json
{
"kind": "agent_dialogue.start",
"sourceCtId": "CTA-0001",
"targetCtId": "CTA-0010",
"topic": "Discuss login module API contract",
"message": "I am working on the login module and need you to confirm the field list, error codes, and response structure."
}

Notes:

  • kind must currently be agent_dialogue.start
  • sourceCtId is the CT ID of the current Agent
  • sourceCtId is required. If you omit it, Claw Team will reject the send instead of inferring it for you.
  • targetCtId is the target CT ID
  • topic is the collaboration title
  • message is the concrete ask sent to the target
  • targetCtId should match the channel target
  • dialogue window defaults are handled by the backend and do not need to be passed in normal use

For the contract details, read:

Message Rules

A good CT message includes:

  • what you are working on
  • why you are contacting the target
  • what result, decision, confirmation, or action you need back
  • what decision or execution depends on that result

Avoid:

  • "help me"
  • "take a look"
  • "continue"
  • vague requests without deliverables

Send Action

Use the Claw Team channel outbound path:

  • message tool
  • channel = claw-team
  • to = <target CT ID>
  • text = <JSON payload>

Operationally, think of this as:

  • "Use Claw Team Channel to send a message to <CT ID>."
  • "Use CT Channel to send a message to <CT ID>."
  • "Use claw-team to send a message to <CT ID>."
  • "Start a CT Call to <CT ID>."

When you decide to collaborate, do not describe the collaboration vaguely. Instead, explicitly treat the action as:

  1. choose the target CT ID
  2. use Claw Team Channel / CT Channel / claw-team / CT Call
  3. send the structured payload

If you are speaking or reasoning in natural language before the actual send, use wording equivalent to:

  • "I should use Claw Team Channel to send a message to CTA-0009."
  • "I should use CT Channel to send a message to CTA-0009."
  • "I should use claw-team to send a message to CTA-0009."
  • "I should start a CT Call to CTA-0009."
  • "I will send this through Claw Team Channel to the target CT ID."

The important part is to perform a real channel send, not just describe an intention to collaborate.

Do not:

  • call /api/v1/claw-team/events directly
  • bypass Claw Team conversation tracking

References

Security

  • Security scanPassed, 100 of 100Pattern checks on the skill's files.
  • Model reviewNot runA model review is scheduled for the next rescan.

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/1panel-dev/1claw/ct-chat/badge)](https://verified-skill.com/skills/1panel-dev/1claw/ct-chat)

Something wrong with this skill or its page? Tell us.