RoninForge.org is an independent open-source workshop that keeps dated, reproducible records of the AI developer tooling economy and builds local-first tools for people who run AI coding agents. It publishes two open datasets under CC BY 4.0: ai-price-index, a dated record of first-party AI model prices with full price history, and State of MCP, a reproducible health and spec-conformance census of every server in the official MCP registry. Its MIT-licensed tools, including the Claude Code spend monitor BudgetClaw and the MCP server health checker Akashi, run entirely on the user's machine and require no API keys. RoninForge is operated day to day by AI agents, with every release reviewed by a single human maintainer.
roninforge.org | About, and the dated facts
Both are open data under CC BY 4.0. Every number carries the date it was measured and the method that produced it, so a third party can reproduce it rather than take our word for it.
A dated, keyless health and spec-conformance census of the official Model Context Protocol server registry. Every server is checked with the same public, unauthenticated probes and given a verdict.
As of the 2026-08-03 census, of 19,804 servers in the registry: 15,974 healthy (80.7%), 2,792 degraded (14.1%), 851 dead (4.3%), 187 unknown (0.9%). Against the 2026-07-28 specification, of the 5,515 servers where readiness could be measured, 6 were spec-ready, 711 at risk, and 4,798 needed migration.
Reproduce any single row with akashi check <server>. Dataset · Methodology · Repository · Atom feed
Cite this dataset: RoninForge.org, "State of MCP: a dated census of the official Model Context Protocol registry", 2026-08-03. CC BY 4.0. https://roninforge.org/data/state-of-mcp/
An open, dated, first-party-sourced record of AI model API prices over time. Most price lists tell you what a model costs today. This one tells you what it cost on a given date, when that price was last confirmed, and links the first-party source that proves it.
Currently 309 prices across 101 models and 11 providers: AI21, Alibaba, Amazon, Anthropic, Cohere, DeepSeek, Google, Meta, Mistral, OpenAI, and xAI.
Dataset · Repository · Hugging Face · DOI 10.5281/zenodo.20730240 · MCP server
MIT licensed. Every one runs on your machine and requires no API keys.
- BudgetClaw by RoninForge - local spend monitor for Claude Code. Tracks cost per project and git branch, warns before a cap, and stops the process on breach. Reads local telemetry only: zero keys, zero prompts, zero latency added. (source)
- RoninForge Akashi - verify a Model Context Protocol server is alive and conformant, keyless. Go CLI and GitHub Action that probes the registry, GitHub, npm, PyPI, and a capability-only MCP initialize handshake, then reports healthy, degraded, or dead with an embeddable verified badge. This is the tool that produces the State of MCP census. (source)
- RoninForge Hanko - validate Claude Code plugin manifests before submission. Go CLI and GitHub Action that catches reserved marketplace names, duplicate hooks, path traversal, and schema errors. (source)
- RoninForge Tsuba - scaffold marketplace-ready Claude Code skills, plugins, hooks, and agents in seconds. Go CLI and GitHub Action. (source)
- claude-code-cost by RoninForge - see your Claude Code spend per project and git branch without leaving Claude Code. A plugin that prices the session logs Claude Code already writes, at the rate that was live on the day each response ran. (source)
- Goei by RoninForge - the Claude Code cost dashboard for your team. Each developer runs the open-source budgetclaw CLI; Goei dedupes across every machine and teammate: cost per project, per developer, per git branch. Zero API keys. (local agent, or npx goei-sync for a zero-install push)
curl -fsSL https://roninforge.org/get | sh # BudgetClaw
curl -fsSL https://roninforge.org/akashi/install.sh | sh
curl -fsSL https://roninforge.org/hanko/install.sh | sh
curl -fsSL https://roninforge.org/tsuba/install.sh | shOr via Homebrew:
brew install roninforge/tap/budgetclaw
brew install roninforge/tap/hanko
brew install roninforge/tap/tsubaNo install, no account, no upload. Everything runs in your browser.
- GitHub Copilot AI Credits calculator - estimate your monthly credit burn under Copilot's usage-based billing from the official per-model token rates, and see whether your plan allowance covers it.
- Copilot usage CSV analyzer - drop the usage report GitHub emails you and see AI credits broken down by model, day, SKU, and user. The file never leaves your browser.
- LLM API pricing comparison - Claude, GPT, Gemini, DeepSeek, Mistral, and Grok token prices side by side, each verified against the official pricing page.
Framework-specific rule packs for Cursor: verified rules, migration skills, and anti-pattern detection so the model generates code that works with current framework versions. Tailwind v4, Next.js 16, Django 5, Rails 8, FastAPI, Go, Spring Boot, ASP.NET Core 9, Laravel 13, Vue 3.5 + Nuxt 4, SvelteKit 2, Astro 6, React Router 7, Expo 55, Kotlin + Compose, Terraform, Drizzle, Playwright, Hono v4, and more.
- How to set a hard daily spend cap on Claude Code
- How to see ongoing project costs across all your Claude Code branches
- How to build and ship a Claude Code plugin end-to-end
Tools MIT licensed, data CC BY 4.0. No telemetry, no lock-in, no accounts.
RoninForge.org is not affiliated with the Ronin blockchain, Hanko.io, or the AKASHI attendance service.