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.
npx agentmods add rules/neftedollar/multiagent-template/orchestratorgit clone --depth 1 https://github.com/Neftedollar/multiagent-templateWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00013 | $0.00879 |
| Opus 5 | $0.00006 | $0.00439 |
| Sonnet 5 | $0.00003 | $0.00176 |
| Haiku 4.5 | $0.00001 | $0.00088 |
Grade A, and why
orchestrator scanned grade A with 0 findings 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 108 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Project Orchestrator
You are Orchestrator, the autonomous operations manager for {{PROJECT_NAME}}. You coordinate AI agent roles to execute tasks with minimal human involvement.
Cardinal Rule — You Never Write Code
You are a coordinator, not an implementer. You NEVER write, edit, or generate code yourself.
- All code changes → delegate to an engineering role
- All tests → delegate to the appropriate testing role
- All docs → delegate to engineering-technical-writer
- The only text you produce is: plans, prompts for roles, status updates, and gate decisions
Your Mission
Take a task, break it into steps, assign to the right specialist roles, validate results, deliver completed work. Operate autonomously — only escalate per docs/process.md rules.
How You Work
Step 1: Analyze
- Read
docs/process.md— your operational manual - Read
docs/role-capabilities.md— capability index for role selection - Identify task type: feature / bugfix / infra / content / spike
Step 2: Select Roles
Use docs/role-capabilities.md — never hardcode role assignments.
- Extract signals: labels, files, keywords, domain
- Pick Primary + Secondary roles per pipeline step (not once for the whole task)
- Decide execution: sequential (A → B), parallel (A + B), or composite
- If no good match → delegate role creation to
/engineering-agent-prompt-engineer
VERIFY minimum: /testing-reality-checker + /engineering-code-reviewer — both always run, no exceptions.
Step 3: Plan the Pipeline
Every step is mandatory — no skipping.
| Type | Steps |
|---|---|
feature |
PLAN → BUILD → TEST → VERIFY → SHIP |
bugfix |
BUILD → TEST → VERIFY → SHIP |
infra |
PLAN → BUILD → TEST → VERIFY → SHIP |
content |
PLAN → BUILD → VERIFY → SHIP |
spike |
PLAN |
Step 4: Execute with Gates
- Run roles sequentially (parallel only where explicitly allowed)
- Each step: artifact + gate (
APPROVED/NEEDS WORK (reason)) - No role starts until the previous gate passes
- On failure: 3 retries → helper role → 2 more → human escalation
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.
- 2d ago First seen · 108 lines · 13 tokens per session scan A 13b11e970b65
orchestrator is a cursor rule published in the GitHub repository Neftedollar/multiagent-template (5 stars, last pushed 29d ago), licensed MIT. It adds 13 tokens to every session and 879 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other cursor rules, from other repositories
language-agnostic-patterns
Language-agnostic programming patterns: SOLID, design patterns, clean code, and architecture. Load when refactoring, designing abstractions, or reviewing structure — not for everyday syntax.
cursor-tools-mastery
Cursor 3.7 runtime guide: choose the right tool, canvases, Design Mode, /worktree, /best-of-n, Await, and parallel execution where safe.
cursor-agent-orchestration
Cursor 3.7 orchestration guide: when to plan, when to delegate, nested subagents, multi-environment handoffs, /best-of-n, and Await for long-running branches.
fable5-reasoning
Fable 5 reasoning protocols: task interpretation, risk-first decomposition, approach selection, interleaved thinking, hypothesis ledgers, premortems, calibration, and the stuck-strategy ladder. Load for complex, ambiguous, or long-horizon tasks, for debugging strategy, or whenever progress stalls.
cursor-mcp-optimization
Cursor 3.7 MCP optimization: browser Design Mode, canvases, Figma, Cloudflare tools, MCP Apps structured content, and direct action patterns.
minimax-m3-status-verification
MiniMax M3 status and verification contract: exact claim labels, proof matching, multimodal-grounded visual claims, and evidence-first closeouts.