Launchpad prepares a repository for AI-assisted development before the agent starts.
It scans the project, resolves engineering standards, generates grounded project context, and emits durable preparation artifacts through AGENTS.md and .ai/*.
Standards-first, not standards-after.
Important
Project status: focused hardening
Launchpad is currently in a focused hardening phase.
The project is being refined around a clearer philosophy: agent-agnostic, preparation-first, standards-aware, and complementary to existing AI development tools.
Launchpad is not trying to replace Claude context mode, Cursor repo context, Windsurf indexing, Copilot workspace features, MCP servers, or IDE-native assistants. Those tools are already good at helping agents inspect a repository during a coding session.
Launchpad's role is different: it prepares durable project context before the session begins, so any agent or context-aware tool starts from a clearer, standards-aligned project contract.
Local AI remains a first-class execution option, but Launchpad is not intended to be limited to local AI only. The long-term direction is to support user choice: deterministic preparation, local AI, paid/cloud providers, hybrid workflows, and agent-specific output adapters.
Launchpad is still in its early stages and currently supports Spring Boot Java projects on Maven or Gradle only.
- 🔍 Scans your project deterministically -- extracts structure, dependencies, endpoints, and documentation without guessing
- 📏 Resolves your engineering standards -- rules, skills, and checklists from a YAML pack you control
- 🧠 Generates grounded context files --
AGENTS.md+.ai/*ready before the agent starts - 📦 Emits durable preparation artifacts -- structured project context, standards, readiness signals, and agent-facing guidance
- 🤝 Complements existing AI tools -- works alongside Claude context mode, Cursor, Windsurf, Copilot, MCP servers, IDE assistants, and local agents
- 🔌 Provider-pluggable -- local AI by default (Ollama, LM Studio, or any OpenAI-compatible endpoint); paid and cloud providers also supported
Launchpad is the preparation layer for AI-assisted development.
Context-aware tools help agents read and reason over a repository during a coding session.
Launchpad prepares the repository before that session begins.
Context mode = runtime awareness
Launchpad = preparation infrastructure
The goal is not to replace agents.
The goal is not to replace context mode.
The goal is to make every agent start smarter.
Launchpad is:
- 🧩 Agent-agnostic -- works with the user's preferred agent, IDE, model, or workflow
- 🛠️ Preparation-first -- makes the repo ready before AI-assisted work begins
- 📐 Standards-aware -- treats engineering rules and checklists as part of the generated project contract
- 🏠 Local-capable -- keeps local AI and deterministic workflows as first-class options
- 🌐 Not local-limited -- leaves room for paid/cloud providers and hybrid workflows when they produce better results
Providers are pluggable. Local AI is the default -- Ollama, LM Studio, llama.cpp, vLLM, or anything with an OpenAI-compatible endpoint -- and keeps your code on your machine, which makes it a first-class option for privacy-sensitive, offline, air-gapped, or cost-sensitive workflows. Paid and cloud providers are also supported when they produce better results: Claude (Anthropic) is the first cloud provider, opt-in and explicitly selected. Local stays the default, and auto-detection never selects a paid provider -- choosing one is always a deliberate choice.
Selecting Anthropic is explicit (launchpad.ai.provider=anthropic); its key reads from LAUNCHPAD_ANTHROPIC_API_KEY (falling back to the shared LAUNCHPAD_LLM_API_KEY). If the key is missing or Anthropic is unreachable, preparation degrades cleanly to deterministic-only output.
Default:
http://localhost:11434
Docs · Getting Started · MIT License · Issues