What does Agent Manager Skill do?

Agent Manager Skill runs multiple local CLI agents in parallel, each in its own tmux session, and provides commands to start/stop agents, tail their logs, assign tasks, and schedule recurring agent work via cron.

What are the system requirements for Agent Manager Skill?

You need tmux and python3 installed locally. It is not applicable if you only need a single agent running interactively.

How do I assign a task to an agent?

Use the `assign` command with an agent identifier and pipe task instructions via stdin, for example: `python3 agent-manager/scripts/main.py assign EMP_0002 <<'EOF' Follow teams/fractalmind-ai-maintenance.md Workflow EOF`

Can I monitor an agent's output in real time?

Yes, use the `monitor` command with the `--follow` flag to tail an agent's logs in real time, for example: `python3 agent-manager/scripts/main.py monitor EMP_0001 --follow`

Skill

Manage and Orchestrate CLI Agents

Name: Agent Manager Skill
Availability: OnlineOnly
Author: Antigravity

CLI tool for running multiple local agents in parallel tmux sessions, assigning tasks, tailing logs, and scheduling recurring work.

Get skill

Works with tmuxpython

Antigravity

Own this? Claim it

Spark score

out of 100

Updated 9 days ago

Version 15.7.0

Add to Favorites

Why it matters

Streamline your command-line operations by managing multiple local CLI agents concurrently. This skill allows for parallel execution, log monitoring, task assignment, and scheduled agent work.

Outcomes

What it gets done

Run multiple CLI agents in separate tmux sessions.

Start, stop, and monitor agent logs in real-time.

Assign tasks to agents and track their output.

Schedule recurring agent tasks using cron.

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/ag-agent-manager-skill | bash

Overview

Agent Manager Skill

A CLI tool for running multiple local agents in parallel tmux sessions, with commands to start, stop, monitor, assign tasks, and schedule recurring work. Use when managing several local CLI agents at once; requires tmux and python3 installed locally.

What it does

Agent Manager Skill runs multiple local CLI agents in parallel, each in its own tmux session, and gives you commands to start/stop agents, tail their logs, assign tasks, and schedule recurring agent work via cron.

python3 agent-manager/scripts/main.py doctor
python3 agent-manager/scripts/main.py list
python3 agent-manager/scripts/main.py start EMP_0001
python3 agent-manager/scripts/main.py monitor EMP_0001 --follow
python3 agent-manager/scripts/main.py assign EMP_0002 <<'EOF'
Follow teams/fractalmind-ai-maintenance.md Workflow
EOF

When to use - and when NOT to

Use this when you need to run multiple local CLI agents in parallel, start/stop them individually, monitor their output in real time, assign them tasks, or schedule recurring agent work via cron.

Not applicable without tmux and python3 installed locally, or if you only need a single agent running interactively - the tmux-session-per-agent model is specifically for managing several agents at once.

Inputs and outputs

Inputs: an agent identifier (e.g. EMP_0001) configured under an agents/ directory, and task instructions to assign (piped in via stdin, e.g. referencing a workflow markdown file).

Outputs: running agents in dedicated tmux sessions, tailed/followed log output via monitor --follow, and assigned tasks executed by the target agent.

Integrations

tmux (session management), python3 (the manager's CLI), cron (for scheduling recurring agent work).

Who it's for

Developers or teams running a fleet of local CLI agents who need a lightweight way to start, stop, monitor, and assign tasks to them without building custom orchestration.

FAQ

Common questions

Discussion