Boomtick is a template repository providing foundational developer tooling, workflow orchestration, and an advanced AI review framework for autonomous agent systems. While framework-agnostic at its core, it includes built-in verification pipelines optimized for Vite applications (as demonstrated in arii/tech-dancer).
Boomtick strictly separates human-led strategic system design from agentic execution—reserving automation for repetitive engineering tasks, compilation, data validation, and operational compliance.
Initialize the agent environment and set up the local CLI:
./setup-agent.shRun the MCP server locally (requires Node.js):
cd mcp
pnpm install
pnpm build
node dist/index.jsThe codebase is split into two primary operational layers:
mcp/(Model Context Protocol Server): Empowers AI agents with structured access to GitHub Pull Requests, repository state, CI logs, and validation tools. Facilitates conflict resolution, branch creation, and interaction with the Jules macro-agent.cli/(td-cli): A terminal-based fallback and local automation toolkit. Integrates directly with the GitHub CLI (gh) to handle manual PR audits, conflict detection, and runtime consistency checks.
Boomtick uses a direct standalone package, container, and composite action integration strategy. See docs/onboarding.md for how to integrate Boomtick into downstream repositories via PyPI (boomtick), Docker (ghcr.io/arii/boomtick), npm (@arii/boomtick-mcp), or composite actions.
- GitHub Models API: Low-latency code reviews and blast-radius analysis via OpenAI-compatible endpoints.
- Gemini API: Multi-modal analysis for deep contextual and visual interface reviews.
- Orchestration: Streamlined LLM routing and fallback strategies powered by
@langchain/core. - Contextual Retrieval: RAG pipeline integration utilizing vector stores for automated triage and overlap analysis.
- Docker Container Packages & PyPI Packaging
- Advanced RAG / Vector Store enhancements
- Developer Onboarding & Workflows — Onboarding guide, codebase context, and collaboration workflows.
- Agent Contracts and Standards — Invariant rules for autonomous agent behavior.
- Agent Workflows & MCP Protocol — Tooling hierarchies (MCP →
td-cli→ Bash) and enforcement rules. - CLI Developer Tooling — Guide to
td-cli, the Tier 2 local developer fallback. - MCP Server Configuration — Setup, capabilities, and tools for the Boomtick MCP.
- Release Process — Guidelines for deployment and release pipelines.
- Impact Analysis Integration — Running and integrating blast-radius checks.
- MCP Testing — Step-by-step verification protocols.
Created by Ariel Anders, creator of Boomtick Blog.
Tags: #DevAI #EngineeringOperations #GitHubModels #GeminiAPI #LangChain #RAG #VectorStore #AgenticExecution #MCP #TypeScript #Python #Vite