Agent Bus

    1

    Coordinate local AI agents with shared task claims, handoffs, and on-demand blackboard checks.

    $1.58

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    agent-bus

    Example session with this skill installed

    Show an example of how two agents should avoid doing the same task with agent-bus. Do not send messages or perform external actions.

    • Read your context and instructions
    • Compiled the agent-bus

    Example coordination flow

    Illustrative example: the following calls and replies explain the workflow; they are not a record of an executed session.

    1. Agent A requests claim(topic="docs:release-notes", by="agent-a").
    2. If the server replies "claimed", Agent A owns that topic and can proceed within the user-authorized scope. A claim does not itself authorize sending, posting, or deploying.
    3. Agent B requests the same topic. An "ALREADY CLAIMED" reply means it should stop duplicate work and inspect the existing result.
    4. When finished, Agent A calls release(topic="docs:release-notes", by="agent-a", result="Release notes prepared for review").
    5. If another agent needs to continue the task, handoff carries the task and context. The receiver checks its inbox when prompted; it is not automatically awakened by a handoff.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Give AI agents a shared coordination routine through the agent-bus MCP server.

    Before an outward action, the agent claims a stable topic so other agents can see who is handling it. Task handoffs carry the context the receiving agent needs. On request, the agent checks its inbox and the shared board, acknowledges work, and records a short result when finished.

    Includes a SKILL.md covering claim and release, handoff, inbox processing, consistent agent identity, and concise board updates. Requires an installed and connected agent-bus MCP server.

    Install by copying the agent-bus folder into ~/.claude/skills/ or ~/.codex/skills/.

    Source: https://github.com/SLYysl/agent-bus/tree/main/skills/agent-bus

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed today

    Software
    agent-bus MCP server

    Permissions

    Network Access

    Uses connected agent-bus MCP tools to read and write coordination records. External actions still require user authorization.

    Listedtoday

    What's inside

    Frequently Asked Questions

    This skill acts as a traffic controller for multiple AI agents, using a shared blackboard to ensure they don't perform the same task twice, such as double-posting to social media or creating conflicting code commits.

    The skill is compatible with any AI agent that supports Model Context Protocol (MCP) tool calls, such as Claude Code or Codex, provided the agent-bus MCP server is active.

    The download contains one SKILL.md instruction file for coordinating agents through agent-bus. The MCP server is installed separately from the linked public repository.

    No, agents do not self-trigger; they require a user prompt or an external watcher to initiate an 'inbox check' to see if new tasks have been assigned on the blackboard.

    If an agent crashes or hangs, the auto-expiration management system identifies stale claims and releases them, ensuring your workflow doesn't remain permanently stalled.

    You must have the agent-bus MCP server pre-installed and running on your local machine for this skill to function, as it provides the underlying infrastructure for the shared blackboard.

    New in Agents & Orchestration