orchestrator-multi

A command for coordinating several coding agents on one software task. It gives agents separate Git worktrees, tracks their progress, and combines their changes when they finish.

In plain words
What is it for?
It is for launching parallel implementation work, checking agent status, and tearing down the worktrees and branches created for the session.
Why use it?
It helps divide a suitable task into independent pieces so multiple agents can work at the same time without sharing one changing folder. It also provides commands to inspect or stop the session.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/intellegix/intellegix-code-agent-toolkit/orchestrator-multi
Clone the repo
git clone --depth 1 https://github.com/intellegix/intellegix-code-agent-toolkit
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 8,610 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.08610
Opus 5 $0.00000 $0.04305
Sonnet 5 $0.00000 $0.01722
Haiku 4.5 $0.00000 $0.00861

Measured 2d ago against content hash 45b881a95f26, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

orchestrator-multi scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Hidden instructionshighPrompt injection

Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.

each agent's CLAUDE.md: `<!-- WARNING: Research step failed — instructions may be incomplete -->`
commands/orchestrator-multi.md · 890 lines

How it starts

The opening of the file, as written. The whole thing — 890 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/orchestrator-multi — Multi-Agent Parallel Orchestration

YOU ARE A MULTI-AGENT ORCHESTRATOR. You split work across N parallel Claude Code agents using git worktrees, write scoped instructions, launch parallel loops, monitor progress, and merge results. You do NOT write implementation code yourself.

Usage

/orchestrator-multi                                        # Auto-discover from cwd + BLUEPRINT.md
/orchestrator-multi <project-path> "<task-description>" [--agents N]  # Explicit mode
/orchestrator-multi status                                 # Show running agent progress
/orchestrator-multi off                                    # Tear down running session

Arguments: $ARGUMENTS

Parse rules (checked in order):

  1. status → show status of running agents (commits, iterations, state)
  2. off → tear down running multi-agent session (remove worktrees, delete branches)
  3. If $ARGUMENTS starts with a path → use that project + remaining text as task
  4. --agents N → number of parallel agents (default: auto, max: 4)
  5. If $ARGUMENTS is empty → Auto-Discovery Mode (Phase 0)

Appropriateness Gate — MANDATORY, RUNS FIRST

Before ANY other phase, evaluate whether this task actually warrants multi-agent parallel execution. If it doesn't, invoke /orchestrator instead. This gate runs after argument parsing but before Phase 0.

Skip this gate (proceed directly to Phase 0) ONLY if:

  • User explicitly passed --agents N with N >= 2
  • User explicitly said "multi-agent", "parallel agents", or "split across agents"
  • Arguments are status or off (subcommands, not new tasks)

Evaluation Criteria

Gather quick context (< 30 seconds, no deep exploration):

  1. git rev-parse --show-toplevel — get project root
  2. git ls-files | wc -l — tracked file count
  3. Check for BLUEPRINT.md — does it have 3+ independent phases?
  4. If task description provided: does it describe work touching 2+ independent modules/directories?

Read the full file on GitHub · 890 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 890 lines · 0 tokens per session scan C 45b881a95f26

Subscribe to this mod's changes

orchestrator-multi is a command published in the GitHub repository intellegix/intellegix-code-agent-toolkit (57 stars, last pushed 8d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 8,610 tokens. A static security scan graded it C with 1 finding (hidden instructions). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.