Skip to content

Releases: coderadius-ai/coderadius

Release list

CodeRadius is open source. 🎉

CodeRadius is open source. 🎉 Latest
Latest

Choose a tag to compare

View all tags
@emnlmn emnlmn released this 06 Jul 12:58
v0.1.0
7f10b55
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

CodeRadius builds a live knowledge graph of your architecture from static analysis: every service, API, queue and database, across all your repos and languages. Any diff can then be
turned into its cross-repo blast radius before merge, and your AI agents read the same graph over MCP.

Install

npm install -g coderadius

or, without a package manager:

curl -fsSL https://coderadius.ai/install | bash

Prebuilt, self-contained binaries for macOS and Linux (x64 and arm64) are attached below. No runtime to install.

What you get

cr blast is like terraform plan for architecture. It computes the blast radius of your changes in seconds and exits with a semantic code (0 safe, 1 watch, 2 breaking), so CI pipelines
and coding agents can branch on the result without parsing text.

cr ui generates the architecture dashboard: service inventory, dependency map, SPOF and coupling analysis, governance violations and dependency health, all in a single self-contained
HTML file you can open locally, share with the team or attach to a PR. This is what it looks like on a demo fleet.

cr mcp start exposes the graph to any MCP-compatible agent (Claude Code, Cursor, Windsurf, Gemini CLI) through ten tools: data contracts, downstream consumers, change impact,
architecture hotspots. The agent checks what it is about to touch before writing code.

Governance is code: declarative YAML policies evaluated against the live graph. Unowned services, deprecated dependencies on exposed APIs, shared-database anti-patterns.

It speaks TypeScript, PHP, Python, Go and Java, with ten frameworks regression-gated by committed eval fixtures. And it is local-first: bring your own LLM key or run fully local with
Ollama. Your code never leaves your machine.

Status

This is a 0.x released early to gather real-world feedback. The deterministic core (AST parsing, taint analysis, graph building, blast radius) is pinned by more than 5,400 tests, while
LLM extraction quality still varies by stack and provider. If it breaks on your codebase, that is exactly the feedback we need: open an
issue.

📖 Docs: https://coderadius.ai/docs

Assets 11
Loading