SuperSpecs: The Discipline Framework That Stops Your AI Agent From Going Rogue. How spec-driven planning, enforced TDD, and a wiki that never forgets finally tame the chaos of AI-assisted software development
SuperSpecs is a markdown-based discipline framework that adds spec-driven planning, TDD and a persistent wiki to AI coding agents.
- for
- Developers using AI coding agents who need repeatable, documented workflows.
- pricing
- open source
- license
- MIT
works withClaude CodeCursorOpenCodeGitHub CopilotCodexGemini CLI
Key features
6 features of SuperSpecs
- Spec-fit for context window — Specs are written to fit a fresh 200k-token window, avoiding reliance on prior chat history.
- TDD without exceptions — Enforces RED-before-GREEN; failing code is deleted until tests pass.
- Persistent wiki memory — After each shipped feature, knowledge is distilled into an interlinked wiki for future sessions.
- Parallel subagent execution — Runs up to five agents in parallel, each with a clean context and the same spec.
- Grill before build — A stress-test interview validates specs against the wiki and techstack before execution.
- Universal agent support — Works with Claude Code, Cursor, OpenCode, Copilot, Codex and Gemini CLI via a single install.
Use cases
- Enforce consistent architecture decisions across multiple AI coding sessions.
- Automatically generate and maintain a project knowledge base that survives agent restarts.
- Run parallel AI subagents on separate spec branches to speed up feature development.
SuperSpecs vs alternatives
 SuperSpecs |  LangChain |  LlamaIndex |  LangGraph |  Mastra | |
|---|---|---|---|---|---|
| Best for | Spec-driven AI development | LLM app development | Data-centric LLM apps | Stateful AI agent pipelines | TypeScript AI agent projects |
| Pricing | Open source | Free | Free | Free | Free |
| DevHunt upvotes | 1 | 0 | 0 | 0 | 0 |
| Launched | Nov 2026 | — | — | — | — |
- SuperSpecs vs LangChain: LangChain is a library for building LLM applications, not a discipline framework for AI coding agents.
- SuperSpecs vs LlamaIndex: LlamaIndex focuses on indexing and querying data for LLMs, whereas SuperSpecs manages spec-driven development workflow.
- SuperSpecs vs LangGraph: LangGraph provides stateful agent orchestration, while SuperSpecs adds TDD, wiki memory and spec enforcement.
- SuperSpecs vs Mastra: Mastra is a TypeScript framework for AI agents; SuperSpecs is language-agnostic and adds a discipline layer on top of any agent.
SuperSpecs FAQ
How does SuperSpecs keep knowledge between AI sessions?+
After each feature is shipped, findings are distilled into a structured wiki that is loaded at the start of the next session.
What happens if code is written before a failing test?+
The code is automatically deleted; the framework enforces RED-before-GREEN so the test suite remains the ground truth.
Can SuperSpecs be used with any AI coding agent?+
Yes, a single install symlinks skills to Claude Code, Cursor, OpenCode, Copilot, Codex and Gemini CLI.
Do specs need to fit within the model’s context window?+
Specs are written to fit a fresh 200k-token window; larger plans are automatically decomposed.
Summarized by DevHunt from superspecs.vercel.app · Oct 16, 2026. Details may change; check the official site.
