Agent skill

Cursor Advanced Composer

by jeremylongshore in jeremylongshore/tons-of-skills-marketplace

Advanced Cursor Composer techniques: agent mode, parallel agents, complex refactoring, and multi-step orchestration.

MITAuto-check passedAgent Workflows

Install Cursor Advanced Composer

skills CLI
$ npx skills add jeremylongshore/tons-of-skills-marketplace --skill cursor-advanced-composer -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install jeremylongshore/tons-of-skills-marketplace cursor-advanced-composer --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/jeremylongshore/tons-of-skills-marketplace.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/.curated/cursor-advanced-composer .claude/skills/cursor-advanced-composer && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
cursor-advanced-composer
GitHub stars
2.8k
Token cost
~2.3k tokens
SKILL.md length
641 words
Files
8 (incl. references)
Skills in repo
3,342
Repo updated
First seen
Licence
MIT

At a glance

Advanced Cursor Composer techniques: agent mode, parallel agents, complex refactoring, and multi-step orchestration.

  • Works in 4 steps: Ask Composer for a plan and… → Constrain it to the approved scope and… → Inspect each diff, run focused… → …
  • Advanced composer
  • SKILL.md covers Overview, Prerequisites, Instructions and Output, plus 10 more sections
  • Calls npm and git

What it does

Cursor Advanced Composer is an agent skill from jeremylongshore/tons-of-skills-marketplace. Advanced Cursor Composer techniques: agent mode, parallel agents, complex refactoring, and multi-step orchestration. Triggers on "advanced composer", "composer patterns", "multi-file generation", "composer refactoring", "agent mode", "parallel agents".

Its SKILL.md is about 2.3k tokens, which your agent loads only when the skill is triggered. The skill folder holds 8 other files, including reference files (for example `references/advanced-patterns.md`, `references/complex-refactoring.md` and `references/composer-strategies.md`). Compatibility notes: Designed for Claude Code

It sits in Agent Workflows, covering Subagents and Refactoring. The repository describes itself as: Model-agnostic agent-skills platform with a harness-free canonical layer, verified adapters, and the ccpi package manager. Explore at tonsofskills.com. The licence is MIT.

When your agent uses it

  • Advanced composer
  • Composer patterns
  • Multi-file generation
  • Composer refactoring

Example prompts

  • “advanced composer”
  • “composer patterns”
  • “multi-file generation”
  • “/cursor-advanced-composer”

Requirements

  • Compatibility (from SKILL.md): Designed for Claude Code
  • Pre-approved tools (allowed-tools): Read, Write, Edit, Bash(cmd:*)

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

  1. Ask Composer for a plan and affected-file list before applying any edits.
  2. Constrain it to the approved scope and provide acceptance checks/non-goals.
  3. Inspect each diff, run focused validation, and split or reject changes that exceed scope.
  4. Commit only reviewed, test-backed changes; retain a revert path.

What it can do on your machine

Read from SKILL.md and the folder at commit cfae287. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • Read
    • Write
    • Edit
    • Bash(cmd:*)

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • npm
    • git

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Links to these hosts (documentation or services it may open):

    • docs.cursor.com
    • cursor.com

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

  • Compatibility

    Designed for Claude Code

    From compatibility in the SKILL.md frontmatter.

Context cost

Cursor Advanced Composer loads about 2.3k tokens when it runs, and up to ~4.1k if it reads all its reference files. Until then it costs about 69 tokens; SKILL.md has 641 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~69
When it runs · the whole SKILL.md, loaded when a task matches
~2.3k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~4.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from jeremylongshore/tons-of-skills-marketplace at commit cfae287, republished under its MIT licence (© jeremylongshore). 641 words, ~2,345 tokens.

Download SKILL.mdSave it as .claude/skills/cursor-advanced-composer/SKILL.md (or your agent's skills folder). This skill also uses 7 other files; get the full folder from GitHub.
name
cursor-advanced-composer
description
Advanced Cursor Composer techniques: agent mode, parallel agents, complex refactoring, and multi-step orchestration. Triggers on "advanced composer", "composer patterns", "multi-file generation", "composer refactoring", "agent mode", "parallel agents".
allowed-tools
Read, Write, Edit, Bash(cmd:*)
compatibility
Designed for Claude Code
version
1.19.0
license
MIT
author
Jeremy Longshore <jeremy@intentsolutions.io>
tags
saas, cursor, cursor-advanced

Cursor Advanced Composer

Overview

Use Composer for bounded multi-file proposals when a human owner can inspect the plan, diff, tests, and rollback before applying changes.

Prerequisites

  • A defined task, affected-file boundary, repository rules, and a clean working tree or recoverable branch.
  • Required test commands and a reviewer for changes that alter architecture, security, or data flow.

Instructions

  1. Ask Composer for a plan and affected-file list before applying any edits.
  2. Constrain it to the approved scope and provide acceptance checks/non-goals.
  3. Inspect each diff, run focused validation, and split or reject changes that exceed scope.
  4. Commit only reviewed, test-backed changes; retain a revert path.

Output

  • A reviewed multi-file change proposal with affected files, validation results, and rollback reference.

Error Handling

ConditionSafe response
Composer proposes unrelated filesReject or split the request; do not apply all.
Output conflicts with rules/testsTreat project controls as authoritative and revise manually or with a narrower prompt.
Sensitive context is attachedStop, remove it, and follow exposure handling.

Examples

Ask Composer to add a feature only in src/feature/ and its existing tests, requiring a plan first. Apply no files until the plan and diff are reviewed; run the focused suite and revert the commit if acceptance fails.

Advanced patterns for Cursor Composer including agent orchestration, complex multi-file refactoring, architecture migrations, and quality-control workflows.

Agent Mode Deep Dive

In Agent mode, Composer autonomously chains tool calls: reading files, searching the codebase, running terminal commands, and writing code. It operates in a loop until the task is complete or it hits the 25-tool-call limit.

Agent Tool Capabilities
ToolWhat it doesApproval needed?
Read fileReads any file in workspaceNo
Search codebaseSemantic + text searchNo
Write fileCreates or overwrites filesShows diff first
Edit fileModifies sections of filesShows diff first
Run terminal commandExecutes shell commandsYes (click Allow)
List directoryBrowses folder structureNo
Controlling Agent Behavior

Give the agent constraints in your prompt:

Migrate all React class components in src/components/ to functional
components with hooks.

Constraints:
- Do NOT modify test files
- Preserve all existing prop types
- Replace lifecycle methods with useEffect equivalents
- Keep the same file names and export structure
- Run `npm test` after each file to verify
Checkpoint Pattern

For large refactors, use explicit checkpoints:

Phase 1: Create the new database schema types in src/types/schema-v2.ts.
         Show me the types before proceeding.

Phase 2: Update the repository layer to use the new types.
         Run tests after each repository file change.

Phase 3: Update the API routes to use the new repository methods.

Phase 4: Update the frontend components to match the new API response shapes.

Stop after each phase and show me a summary of changes.

The agent pauses after 25 tool calls automatically. Click "Continue" to allow more.

Parallel Agent Workflows

Run up to 8 Composer agents simultaneously, each in its own tab.

Use Case: Parallel Feature Development
# Agent 1 (Tab: "Auth")
Implement JWT authentication:
- src/middleware/auth.ts
- src/services/token.service.ts
- src/api/auth/login/route.ts
- src/api/auth/refresh/route.ts

# Agent 2 (Tab: "Products")
Implement product catalog:
- prisma/schema additions
- src/api/products/route.ts
- src/services/product.service.ts

# Agent 3 (Tab: "Tests")
Write integration tests for the existing user API:
- tests/integration/users.test.ts

Each agent has its own context and conversation history. They can modify the same codebase simultaneously, but be careful with overlapping files.

Show full SKILL.md (252 more words)Show less
Conflict Resolution Between Agents

If two agents modify the same file, the second write overwrites the first. Mitigate this by:

  • Assigning non-overlapping file scopes to each agent
  • Running agents sequentially for shared files
  • Using git to cherry-pick the best changes

Complex Refactoring Patterns

Pattern: Extract and Replace

Extract a pattern into a shared utility, then replace all occurrences:

@src/api/

I see repeated error handling in every API route:
try { ... } catch (err) { if (err instanceof ZodError) ... }

Step 1: Create src/utils/api-handler.ts with a withValidation()
wrapper that handles the try/catch + Zod pattern.

Step 2: Update ALL route files in src/api/ to use withValidation()
instead of manual try/catch.

Show me api-handler.ts first, then refactor the routes.
Pattern: Interface-First Migration

Define the target interface, then migrate implementations:

@src/services/

Step 1: Create src/interfaces/repository.ts with a generic
Repository<T> interface:
  - findById(id: string): Promise<T | null>
  - findMany(filter: Filter<T>): Promise<T[]>
  - create(data: CreateInput<T>): Promise<T>
  - update(id: string, data: UpdateInput<T>): Promise<T>
  - delete(id: string): Promise<void>

Step 2: Refactor UserService to implement Repository<User>.
Step 3: Refactor ProductService to implement Repository<Product>.
Step 4: Update API routes to use the Repository interface type.
Pattern: Test-Driven Refactoring

Write tests first, then refactor with confidence:

@src/services/payment.service.ts

Step 1: Write comprehensive tests for the current PaymentService
behavior. Cover all public methods and edge cases.
Run `npm test` to verify they pass.

Step 2: Refactor PaymentService to use the Strategy pattern for
different payment providers (Stripe, PayPal, manual).
Run `npm test` after each change to ensure nothing breaks.

Multi-Codebase Orchestration

For monorepos or multi-root workspaces:

@packages/shared/src/types/
@packages/api/src/routes/
@packages/web/src/hooks/

Add a "notifications" feature across all packages:
1. shared: NotificationType enum, Notification interface
2. api: GET /notifications, POST /notifications/mark-read endpoints
3. web: useNotifications hook, NotificationBell component

Use the shared types in both api and web packages.
Import from @myorg/shared (the workspace alias).

Quality Control Workflows

Pre-Apply Review Checklist

Before clicking "Apply All":

  1. Read every diff -- do not blindly apply multi-file changes
  2. Check imports -- Composer sometimes generates wrong import paths
  3. Verify types -- ensure type annotations are correct, not any
  4. Look for hallucinations -- methods or modules that do not exist
  5. Run build -- `Cmd+`` > run your build command after applying
Post-Apply Validation
# In Composer or terminal:
npm run build       # Catches import errors, type mismatches
npm run lint        # Catches style violations
npm run test        # Catches behavior regressions
Rollback Strategy

If applied changes break the build:

bash
# Undo all uncommitted changes
git checkout .

# Or selectively undo specific files
git checkout -- src/api/products/route.ts

Always commit working code before starting a Composer session.

Advanced Prompting Techniques

System-Level Instructions via Rules

Create a Composer-specific rule:

yaml
# .cursor/rules/composer-standards.mdc
---
description: "Standards for Composer-generated code"
globs: ""
alwaysApply: true
---
When generating code via Composer:
- Always include JSDoc comments on exported functions
- Always add error handling (never let functions throw unhandled)
- Generate corresponding test files for new modules
- Use named exports, never default exports
- Import types with `import type` syntax
Iterative Refinement

After first pass:

The generated product.service.ts looks good but:
1. Add pagination support to findMany (page, limit params)
2. Add a findByCategory method
3. Use transactions for createWithItems

Composer retains conversation context, so follow-up instructions build on previous output.

Enterprise Considerations

  • Compliance: Review generated code for license compliance before committing
  • Security: Composer may generate code with SQL injection or XSS vulnerabilities -- always audit
  • Cost: Agent mode with reasoning models (o1, Opus) can consume thousands of tokens per session
  • Audit trail: Commit Composer output with descriptive messages for code review traceability

Resources

© jeremylongshore, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 7 other files (references) in skills/.curated/cursor-advanced-composer of jeremylongshore/tons-of-skills-marketplace.

  • SKILL.md
  • references/advanced-patterns.md
  • references/complex-refactoring.md
  • references/composer-strategies.md
  • references/errors.md
  • references/examples.md
  • references/multi-file-orchestration.md
  • references/quality-control.md

Open the folder on GitHubat commit cfae287

Compare with similar skills

Cursor Advanced Composer next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Cursor Advanced Composer compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Cursor Advanced Composer this skilljeremylongshore/tons-of-skills-marketplace2.8k—~2.3kAutomated safety check: PassMIT
Antigravity Agentsmarkfulton/claude-antigravity-agents130—~2.1kAutomated safety check: PassMIT
Subagent Coordinatorflyxl/datazen114—~936Automated safety check: PassGPL-3.0
Orchestrate Batch RefactorDimillian/Skills4k—~889Automated safety check: PassMIT
Batch Orchestrationrohitg00/pro-workflow2.9k—~1.2kAutomated safety check: PassNone
Codex CLIkortix-ai/suna20k—~1.6kAutomated safety check: PassCustom licence

Similar skills

  • Antigravity Agents

    markfulton/claude-antigravity-agents

    Delegate coding, code review, analysis, and research jobs to Google Antigravity CLI (agy) sub-agents that run alongside your own work.

    130 GitHub stars~2.1k tokensUpdated 2 mo ago
    Agent WorkflowsAuto-check passed
  • Orchestrate multi-track parallel feature development with subagents and git worktrees.

    114 GitHub stars~936 tokensUpdated yesterday
    Agent WorkflowsAuto-check passed
  • Plan and execute large refactor or rewrite efforts efficiently with parallel multi-agent analysis and implementation.

    4k GitHub stars~889 tokensUpdated 6 mo ago
    Agent WorkflowsAuto-check passed
  • Batch Orchestration

    rohitg00/pro-workflow

    Decompose large-scale changes into independent units and spawn parallel agents in isolated worktrees.

    2.9k GitHub stars~1.2k tokensUpdated 12 days ago
    Agent WorkflowsAuto-check passed
  • Codex CLI

    kortix-ai/suna

    Drive OpenAI's Codex CLI (codex exec) as a non-interactive coding sub-agent from inside Claude Code.

    20k GitHub stars~1.6k tokensUpdated today
    Agent WorkflowsAuto-check passed
  • Batch Refactor With Sub Agents

    r3bl-org/r3bl-open-core

    Use a sub-agent (like generalist) to perform repetitive code transformations across multiple files in a single turn.

    485 GitHub stars~816 tokensUpdated yesterday
    Agent WorkflowsAuto-check passed

More from jeremylongshore/tons-of-skills-marketplace

All 3,342 skills in this repo
  • Performing Security Code Review

    jeremylongshore/tons-of-skills-marketplace

    Execute this skill enables AI assistant to conduct a security-focused code review using the security-agent plugin.

    2.8k GitHub starsUsed in 2 repos~1.3k tokens
    Auto-check: notes
  • Adapting Transfer Learning Models

    jeremylongshore/tons-of-skills-marketplace

    Build this skill automates the adaptation of pre-trained machine learning models using transfer learning techniques.

    2.8k GitHub stars~1.1k tokensUpdated today
    Auto-check passed
  • Agent Context Loader

    jeremylongshore/tons-of-skills-marketplace

    Execute proactive auto-loading: automatically detects and loads agents.md files.

    2.8k GitHub stars~1.1k tokensUpdated today
    Auto-check passed
  • Aggregating Performance Metrics

    jeremylongshore/tons-of-skills-marketplace

    Aggregate and centralize performance metrics from applications, systems, databases, caches, and services.

    2.8k GitHub stars~1.2k tokensUpdated today
    Auto-check passed
  • Analyzing Capacity Planning

    jeremylongshore/tons-of-skills-marketplace

    Execute this skill enables AI assistant to analyze capacity requirements and plan for future growth.

    2.8k GitHub stars~947 tokensUpdated today
    Auto-check passed
  • Analyzing Database Indexes

    jeremylongshore/tons-of-skills-marketplace

    Process use when you need to work with database indexing. An agent skill from jeremylongshore/tons-of-skills-marketplace.

    2.8k GitHub stars~2k tokensUpdated today
    Auto-check passed

Questions about Cursor Advanced Composer

What does Cursor Advanced Composer do?

Advanced Cursor Composer techniques: agent mode, parallel agents, complex refactoring, and multi-step orchestration. Cursor Advanced Composer is an agent skill from jeremylongshore/tons-of-skills-marketplace. Advanced Cursor Composer techniques: agent mode, parallel agents, complex refactoring, and multi-step orchestration.

When should I use Cursor Advanced Composer?

Cursor Advanced Composer fits situations like: advanced composer; composer patterns; multi-file generation; composer refactoring.

How do I install Cursor Advanced Composer in Claude Code?

Run `npx skills add jeremylongshore/tons-of-skills-marketplace --skill cursor-advanced-composer -a claude-code`. Or copy the skill folder (skills/.curated/cursor-advanced-composer in jeremylongshore/tons-of-skills-marketplace) into .claude/skills/cursor-advanced-composer in your project. Claude Code loads it when a task matches its description.

How do I install Cursor Advanced Composer in Codex?

Run `npx skills add jeremylongshore/tons-of-skills-marketplace --skill cursor-advanced-composer -a codex`. Or copy the skill folder (skills/.curated/cursor-advanced-composer in jeremylongshore/tons-of-skills-marketplace) into .agents/skills/cursor-advanced-composer in your project. Codex loads it when a task matches its description.

Can I use Cursor Advanced Composer in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add jeremylongshore/tons-of-skills-marketplace --skill cursor-advanced-composer -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/cursor-advanced-composer, .gemini/skills/cursor-advanced-composer, .github/skills/cursor-advanced-composer and .opencode/skills/cursor-advanced-composer in your project.

What does Cursor Advanced Composer need to run?

Going by SKILL.md and its folder, Cursor Advanced Composer needs the command-line tools its instructions call (npm and git). Its frontmatter pre-approves these tools: Read, Write, Edit, Bash(cmd:*). Compatibility (from SKILL.md): Designed for Claude Code.

Does Cursor Advanced Composer access the network?

SKILL.md names 2 domains. As links in the text: docs.cursor.com and cursor.com. This is read from the text; nothing was executed.

Is Cursor Advanced Composer safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Cursor Advanced Composer use?

Cursor Advanced Composer is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Cursor Advanced Composer use?

About 2.3k tokens (SKILL.md is roughly 9.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 1.8k tokens, read only when the agent opens those files.

What are the alternatives to Cursor Advanced Composer?

Skills that share tags, products or a category with Cursor Advanced Composer: Antigravity Agents (markfulton/claude-antigravity-agents, 130 stars), Subagent Coordinator (flyxl/datazen, 114 stars), Orchestrate Batch Refactor (Dimillian/Skills, 4k stars) and Batch Orchestration (rohitg00/pro-workflow, 2.9k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Cursor Advanced Composer?

jeremylongshore (a GitHub user) maintains it in jeremylongshore/tons-of-skills-marketplace, which has 2,827 GitHub stars. The repository holds 3,342 skills in this directory. The repository was last updated on October 10, 2026.

Source: jeremylongshore/tons-of-skills-marketplace on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.