plugins/ai-maestro/skills/ai-maestro-agents-management/SKILL.md
Creates, manages, and orchestrates AI agents using the AI Maestro CLI. Use when the user asks to "create agent", "list agents", "delete agent", "rename agent", "hibernate agent", "wake agent", "install plugin", "show agent", "export agent", "restart agent", "install marketplace", or any agent lifecycle management task.
npx skillsauth add 23blocks-os/ai-maestro-plugins ai-maestro-agents-managementInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
4 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Manage AI agents through the AI Maestro CLI. This skill provides commands for creating, updating, deleting, hibernating, and waking agents. It also handles plugin management and agent import/export.
This skill is for managing other agents, not for inter-agent communication (use agent-messaging skill for that).
Script: aimaestro-agent.sh (Bash, macOS/Linux)
Installation: ./install-agent-cli.sh
Requirements: macOS or Linux, Bash 4.0+, tmux 3.0+, jq, curl
aimaestro-agent.sh list [--status online|offline|hibernated|all] [--format table|json|names] [-q|--quiet] [--json]
Examples: list, list --status online, list --format json, list -q
aimaestro-agent.sh show <agent> [--format pretty|json]
--dir is required.
aimaestro-agent.sh create <name> --dir <path> [options] [-- <program-args>...]
Options: -p/--program, -m/--model, -t/--task, --tags, -T/--trust-level, --no-session, --no-folder, --force-folder
Trust levels: supervised (default), planOnly, trustEdits, smartAuto, fullAutonomy
Examples:
aimaestro-agent.sh create my-api --dir /Users/dev/projects/my-api
aimaestro-agent.sh create backend-service \
--dir /Users/dev/projects/backend \
--task "Implement user authentication with JWT" \
--tags "api,auth,security"
aimaestro-agent.sh create debug-agent --dir /Users/dev/projects/debug -- --verbose --debug
aimaestro-agent.sh create auto-agent --dir ~/Code/auto --trust-level smartAuto
aimaestro-agent.sh update <agent> [options]
Options: -t/--task, -m/--model, --tags, --add-tag, --remove-tag, --args
Examples:
aimaestro-agent.sh update backend-api --task "Focus on payment integration"
aimaestro-agent.sh update backend-api --add-tag "critical"
aimaestro-agent.sh update backend-api --args "--continue --chrome"
Destructive operation. Requires --confirm.
aimaestro-agent.sh delete <agent> --confirm [--keep-folder] [--keep-data]
aimaestro-agent.sh rename <old-name> <new-name> [--rename-session] [--rename-folder] [-y]
aimaestro-agent.sh hibernate <agent>
aimaestro-agent.sh wake <agent> [--attach] [-T|--trust-level <level>]
The -T/--trust-level flag overrides the agent's stored trust level for this session only.
Example:
aimaestro-agent.sh wake researcher --trust-level planOnly
aimaestro-agent.sh restart <agent> [--wait <seconds>]
Hibernates, waits (default 3s), then wakes. Cannot restart the current session.
aimaestro-agent.sh session add <agent> [--role <role>]
aimaestro-agent.sh session remove <agent> [--index <n>] [--all]
aimaestro-agent.sh session exec <agent> <command...>
aimaestro-agent.sh plugin install <agent> <plugin> [-s|--scope user|project|local] [--no-restart]
aimaestro-agent.sh plugin uninstall <agent> <plugin> [-s|--scope user|project|local] [--force|-f]
aimaestro-agent.sh plugin update <agent> <plugin> [-s|--scope user|project|local]
aimaestro-agent.sh plugin load <agent> <path> [<path>...]
aimaestro-agent.sh plugin list <agent>
aimaestro-agent.sh plugin enable <agent> <plugin> [-s|--scope user|project|local]
aimaestro-agent.sh plugin disable <agent> <plugin> [-s|--scope user|project|local]
aimaestro-agent.sh plugin validate <agent> <plugin-path>
aimaestro-agent.sh plugin reinstall <agent> <plugin> [-s|--scope user|project|local]
aimaestro-agent.sh plugin clean <agent> [--dry-run|-n]
aimaestro-agent.sh plugin marketplace list <agent>
aimaestro-agent.sh plugin marketplace add <agent> <source> [--no-restart]
aimaestro-agent.sh plugin marketplace remove <agent> <name> [--force|-f]
aimaestro-agent.sh plugin marketplace update <agent> [<name>]
Source formats: owner/repo, github:owner/repo, HTTPS/SSH Git URLs, #branch, local directory, remote URL.
Examples:
aimaestro-agent.sh plugin marketplace add backend-api owner/repo
aimaestro-agent.sh plugin marketplace add backend-api https://github.com/o/r.git#v1.0.0
aimaestro-agent.sh plugin marketplace remove backend-api my-marketplace --force
aimaestro-agent.sh export <agent> [-o <output-file>]
Default output: <agent>.agent.json. Currently exports configuration only.
aimaestro-agent.sh import <file> [--name <new-name>] [--dir <new-dir>]
aimaestro-agent.sh skill list <agent>
aimaestro-agent.sh skill add <agent> <skill-id> [--type marketplace|custom] [--path <path>]
aimaestro-agent.sh skill remove <agent> <skill-id>
aimaestro-agent.sh skill install <agent> <source> [-s|--scope user|project|local] [--name <name>]
Examples:
aimaestro-agent.sh skill install my-agent ./my-skill.skill
aimaestro-agent.sh skill install my-agent ./path/to/skill-folder --scope project
aimaestro-agent.sh skill install backend-api ./debug-skill --scope local
aimaestro-agent.sh skill uninstall <agent> <skill-name> [-s|--scope user|project|local]
Agent not found: aimaestro-agent.sh list to see available agents.
Script not found: Check which aimaestro-agent.sh and verify ~/.local/bin is in PATH.
API not running: curl http://localhost:23000/api/hosts/identity — start AI Maestro if down.
For detailed output formats, scenarios, troubleshooting, error table, and architecture, see references/REFERENCE.md.
tools
Create, manage, and interact with canvas HTML pages. Write visual UIs that users see in the AI Maestro dashboard, receive structured interactions when users click/submit/select, and update pages in response. Full lifecycle management for agent-rendered canvases.
tools
Creates, manages, and orchestrates AI agents using the AI Maestro CLI. Use when the user asks to "create agent", "list agents", "delete agent", "rename agent", "hibernate agent", "wake agent", "install plugin", "show agent", "export agent", "restart agent", "install marketplace", or any agent lifecycle management task.
documentation
Authenticate AI agents with auth servers using the Agent Identity (AID) protocol. Supports Ed25519 identity documents, proof of possession, OAuth 2.0 token exchange, and scoped JWT tokens. Self-contained — works independently without other protocols.
tools
Creates and manages persistent markdown planning files (task_plan.md, findings.md, progress.md) for complex task execution. Use when starting multi-step tasks, research projects, or any task requiring >5 tool calls. Solves the EXECUTION problem - staying focused during long-running tasks.