swarm

swarm is a cursor rule for coding agents from dasomel/oh-my-cursor. It costs 17 tokens per session (761 once invoked), scanned A, original, MIT.

A set of rules for breaking a large coding request into 5 to 10 small, independent tasks. Each task is limited to a file or function and includes a way to check completion.

In plain words
What is it for?
Decomposing large features, multi-file refactors, bulk updates, and repeated test or endpoint changes into focused work units.
Why use it?
It makes broad changes easier to organize and review by separating work that can be completed independently and ordering tasks around their dependencies.

Cursor rule

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 rules/dasomel/oh-my-cursor/swarm
Clone the repo
git clone --depth 1 https://github.com/dasomel/oh-my-cursor

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for swarm

README.md
[![agentmods](https://agentmods.dev/badge/rules/dasomel/oh-my-cursor/swarm.svg)](https://agentmods.dev/rules/dasomel/oh-my-cursor/swarm)
Your own site
<a href="https://agentmods.dev/rules/dasomel/oh-my-cursor/swarm"><img src="https://agentmods.dev/badge/rules/dasomel/oh-my-cursor/swarm.svg" alt="Measured on agentmods" height="20"></a>
Per session 17 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 761 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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 $0.00017 $0.00761
Opus 5 $0.00009 $0.00380
Sonnet 5 $0.00003 $0.00152
Haiku 4.5 $0.00002 $0.00076

Measured 3d ago against content hash 14731a411da6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

swarm 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 3d 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.

rules/workflows/swarm.mdc · 108 lines

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.

Swarm Workflow

Swarm decomposes large requests into 5-10 independent micro-tasks for systematic, focused execution. Inspired by parallel swarm patterns, adapted for Cursor's single-agent model.

Triggers

  • "swarm", "decompose", "break this down", "micro-tasks", "divide and conquer"

When to Use

  • Large features spanning multiple files
  • Refactoring across many modules
  • Bulk operations (add tests to 10 files, update all endpoints, etc.)
  • Tasks where independent pieces can be clearly identified

Protocol

Phase 1: Decompose

Break the request into 5-10 independent micro-tasks. Each micro-task must be:

- Scoped to a single file or function
- Independently completable (no dependencies on other micro-tasks)
- Small enough to complete in one focused session
- Clearly defined with specific deliverables

Task Template

For each micro-task, define:

## Task {N}: {title}
- **File**: {specific file path}
- **Goal**: {atomic, measurable objective}
- **Acceptance**: {how to verify completion}
- **Constraints**: {what NOT to change}

Phase 2: Sequence

Order tasks by dependency:

1. Independent tasks first (can run in any order)
2. Tasks with dependencies after their prerequisites
3. Integration/verification tasks last

Write the task list to .omc-cursor/plans/swarm-{name}.md.

Phase 3: Execute

For each micro-task:

Switching to Executor role.
1. Read ONLY the files relevant to this task
2. Make the minimum change needed
3. Self-verify:
   - Does the change compile/lint?
   - Does it satisfy the acceptance criteria?
   - Did it avoid touching files outside scope?
4. Mark task complete in the plan
5. Move to next task

Phase 4: Verify

After all tasks complete:

Switching to Verifier role.
1. Check cross-file consistency
2. Run full test suite
3. Run build
4. Verify no integration issues between micro-tasks
5. Fix any issues inline (don't re-decompose)

Code Quality Rules (per micro-task)

  • Functions ≤ 20 lines
  • Nesting ≤ 2 levels
  • No ternary chains
  • No complex one-liners
  • Each function does one thing

Read the full file on GitHub · 108 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. 3d ago First seen · 108 lines · 17 tokens per session scan A 14731a411da6

Subscribe to this mod's changes

swarm is a cursor rule published in the GitHub repository dasomel/oh-my-cursor (2 stars, last pushed 5mo ago), licensed MIT. It adds 17 tokens to every session and 761 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.