Agent Manager Skill
Run several local Claude/Codex-style CLI agents in parallel inside tmux, assign tasks, tail logs, and wire recurring jobs with cron.
Overview
Agent Manager Skill is an agent skill for the Build phase that runs and supervises multiple local CLI agents in tmux with start, stop, monitor, assign, and cron-friendly scheduling.
Install
npx skills add https://github.com/davila7/claude-code-templates --skill agent-manager-skillWhat is this skill?
- Start, stop, and list multiple agents as separate tmux sessions
- Monitor and follow agent logs per employee/session ID (EMP_xxxx)
- Assign tasks via stdin heredoc to running agents
- `doctor` and `list` commands for workspace health checks
- Cron-friendly scheduling for recurring agent workloads
- Requires tmux and python3
- Common flow: doctor, list, start, monitor --follow, assign
Adoption & trust: 536 installs on skills.sh; 27.8k GitHub stars; 2/3 security scanners passed (skills.sh audits).
What problem does it solve?
You want several coding agents working in parallel locally but one terminal cannot safely supervise, assign, and tail all of their output.
Who is it for?
Solo builders on macOS or Linux who already use tmux and want a lightweight ops layer over multiple local CLI agents during active development.
Skip if: Teams that only use a single cloud-hosted agent UI with no local tmux/Python setup, or builders who need managed cloud orchestration instead of on-machine sessions.
When should I use this skill?
You need to run multiple local CLI agents in parallel in separate tmux sessions, start/stop and tail logs, assign tasks, or schedule recurring agent work with cron.
What do I get? / Deliverables
You get named tmux-backed agent sessions with CLI control for lifecycle, log follow, task assignment, and repeatable cron runs from your workspace.
- Running tmux sessions per configured agent ID
- Monitored log streams and stdin-assigned task payloads to agents
Recommended Skills
Journey fit
Orchestrating multiple coding agents is core build-time agent tooling—not a ship or grow concern—so the canonical shelf is build → agent-tooling. The skill is explicitly for managing CLI agent sessions, start/stop, monitor, and assign work—exactly the agent-tooling lane for solo builders shipping with local agents.
How it compares
Use for tmux-based local agent fleet control instead of manually spawning detached shells or unrelated process managers with no assign/monitor workflow.
Common Questions / FAQ
Who is agent-manager-skill for?
Indie and solo developers running multiple local CLI agents (Claude Code, Codex, or similar) who want tmux sessions, log monitoring, and scripted task assignment without a full orchestration platform.
When should I use agent-manager-skill?
During the build phase when you need parallel agents for integrations or agent-tooling work—for example starting EMP_0001 and EMP_0002, following logs while you code, or scheduling recurring maintenance via cron.
Is agent-manager-skill safe to install?
It requires cloning a third-party repo and running Python against your filesystem and tmux; review the Security Audits panel on this Prism page and inspect `agents/` configs before assigning broad workflows.
SKILL.md
READMESKILL.md - Agent Manager Skill
# Agent Manager Skill ## When to use Use this skill when you need to: - run multiple local CLI agents in parallel (separate tmux sessions) - start/stop agents and tail their logs - assign tasks to agents and monitor output - schedule recurring agent work (cron) ## Prerequisites Install `agent-manager-skill` in your workspace: ```bash git clone https://github.com/fractalmind-ai/agent-manager-skill.git ``` ## Common commands ```bash 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 ``` ## Notes - Requires `tmux` and `python3`. - Agents are configured under an `agents/` directory (see the repo for examples).