superfast

Superfast

Why Superfast?

From Individual Speed to Enterprise Institutional Wisdom.

The next frontier of AI isn't just about faster models; it's about Enterprise Readiness. While individual agents are productive, they often struggle with the complexity, scale, and persistence requirements of professional engineering environments. Without a shared cognitive map, agents remain transient—repeating work, missing architectural dependencies, and eventually losing the "source of truth."

Superfast is the enterprise-grade evolution of the foundational Superpowers framework. We’ve combined world-class systematic methodology with a high-performance FastMemory engine to turn transient agent sessions into persistent institutional assets.

• Enterprise Data Integration: Bridge the gap between coding agents and massive enterprise datalakes (Microsoft Fabric, AWS Glue, Databricks).
• Cognitive Scaling: Navigate millions of lines of code and complex topologies through a structured CBFDAE logic layer.
• Systematic Rigor: Inherit the strict TDD, Socratic brainstorming, and verifiable planning that made Superpowers great, now scaled for the Fortune 500.

Don't just build features; build a memory moat.

Why FastMemory vs Standard Superpowers?

Superpowers is an incredible framework, but it operates on a critical assumption: that your codebase fits neatly into native AI context windows. That is only true for 80% of low-value, boilerplate codebases.

The remaining 20% of high-value, enterprise codebases are simply too massive and complex to rely on inherent AI memory. Standard agents lose context, hallucinate logic, and break down under scale. That's where FastMemory comes into play. It provides Superpowers with the required nitro boost—mapping large-scale code and content memory into our deterministic CBFDAE ontology, giving your agent perfect institutional wisdom in milliseconds.

[!TIP] Verified SOTA Engine: Built directly on the native FastMemory engine, Superfast harnesses the identical graph architecture that recently achieved State-of-the-Art (SOTA) on 13 core benchmarks over on Hugging Face (including FinanceBench, FRAMES, and GraphRAG-Bench). You are utilizing the most accurate multi-hop reasoning model natively inside your codebase.

How it works

It starts by establishing your Context Memory. Before any brainstorming or planning occurs, it instantly triggers the FastMemory Build-Up Skill, recursively mapping your enterprise repository into a deterministic CBFDAE graph.

Once your agent's cognitive map is established, it steps back and asks you what you're really trying to do.

Once it's teased a spec out of the conversation, it shows it to you in chunks short enough to actually read and digest.

After you've signed off on the design, your agent puts together an implementation plan that's clear enough for an enthusiastic junior engineer with poor taste, no judgement, no project context, and an aversion to testing to follow. It emphasizes true red/green TDD, YAGNI (You Aren't Gonna Need It), and DRY.

Next up, once you say "go", it launches a subagent-driven-development process, having agents work through each engineering task, inspecting and reviewing their work, and continuing forward. It's not uncommon for Claude to be able to work autonomously for a couple hours at a time without deviating from the plan you put together.

There's a bunch more to it, but that's the core of the system. And because the skills trigger automatically, you don't need to do anything special. Your coding agent just has Superfast.

Installation

Note: Installation differs by platform. Claude Code or Cursor have built-in plugin marketplaces. Codex and OpenCode require manual setup.

Claude Code Official Marketplace

Superfast is available via our official GitHub Plugin Registry.

In Claude Code, register the marketplace first:

/plugin marketplace add https://raw.githubusercontent.com/FastBuilderAI/superfast/main/.claude-plugin/marketplace.json

Then install the plugin directly:

/plugin install superfast

Cursor (via Plugin Marketplace)

In Cursor Agent chat, install from marketplace:

/add-plugin superfast

or search for "superfast" in the plugin marketplace.

Codex

Tell Codex:

Fetch and follow instructions from https://raw.githubusercontent.com/obra/superfast/refs/heads/main/.codex/INSTALL.md

Detailed docs: docs/README.codex.md

OpenCode

Tell OpenCode:

Fetch and follow instructions from https://raw.githubusercontent.com/obra/superfast/refs/heads/main/.opencode/INSTALL.md

Detailed docs: docs/README.opencode.md

Gemini CLI

gemini extensions install https://github.com/FastBuilderAI/superfast

To update: