orchestrate

name	orchestrate
description	Plan and execute multi-agent workflows using lionagi's CLI: li o flow (DAG pipelines), li o fanout (parallel workers), and li play (playbook invocations). Use when a task needs multiple agents working in parallel or staged phases.
allowed-tools	["Bash","Read","Write","Glob","Grep"]

orchestrate

Plan and execute multi-agent workflows using lionagi's CLI.

When to use which command

Situation	Command
Single task, one agent	li agent MODEL PROMPT
Same prompt to N independent workers	li o fanout MODEL PROMPT -n N
Staged pipeline with dependencies	li o flow MODEL PROMPT
Pre-saved parametric workflow	li play NAME [PROMPT]

If all subtasks are independent (no output feeds another), use fanout. If any subtask depends on the output of another, use flow. li play NAME is sugar for li o flow -p NAME.

Quick start

# Single agent
li agent claude "Write unit tests for auth.py"

# Fan out 4 parallel workers + synthesize
li o fanout claude "Review this codebase for security issues" -n 4 \
    --with-synthesis --save ./out --yolo --bypass

# DAG flow — dry-run first, then execute
li o flow claude "Audit auth, implement fixes, verify with tests" \
    --dry-run --effort high
li o flow claude "Audit auth, implement fixes, verify with tests" \
    --with-synthesis --save ./flow-out --yolo --bypass

# Run a saved playbook
li play security-audit "JWT middleware" --save ./out
li play list  # list available playbooks

Key principles

• Critic runs last — never parallel with producers. Set control=True.
• Agent reuse > spawning — reusing agent_id across ops preserves memory.
• Artifact handoff — agents write to {save_dir}/{agent_id}/, downstream reads from ../{dep_id}/.
• depends_on is mandatory for every non-root op.
• --dry-run before executing — preview the DAG before committing.

Companion references

For detailed documentation, read these companion files in this skill directory:

• cli-reference.md — complete flag tables for li agent, li o fanout, li o flow, li play, li team, li invoke
• dag-planning.md — FlowPlan data model, DAG decomposition principles, role-to-model routing, re-plan rounds
• workflows.md — standard workflow patterns (parallel exploration, staged pipeline, background flow, spec files, visualization)
• teams-and-tracking.md — team coordination patterns, invocation tracking, scheduling

Source code

Component	Path
CLI entrypoint	lionagi/cli/main.py
Flow engine (FlowPlan, FlowOp, FlowAgent)	lionagi/cli/orchestrate/flow.py
Fanout engine	lionagi/cli/orchestrate/fanout.py
Argparse definitions	lionagi/cli/orchestrate/__init__.py
Agent CLI	lionagi/cli/agent.py
Teams	lionagi/cli/team.py
Invocations	lionagi/cli/invoke.py
Scheduler engine	apps/studio/server/scheduler/engine.py