Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/coderadius-ai/coderadius/claude-mdgit clone --depth 1 https://github.com/coderadius-ai/coderadiusWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.05795 | $0.05795 |
| Opus 5 | $0.02898 | $0.02898 |
| Sonnet 5 | $0.01159 | $0.01159 |
| Haiku 4.5 | $0.00579 | $0.00579 |
Grade A, and why
coderadius CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
This is a copy
98% identical to coderadius AGENTS.md — 4 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 263 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
What This Is
CodeRadius is a CLI tool that builds an architectural knowledge graph from polyglot codebases. It statically analyzes source code (TypeScript, PHP, Go, Python), extracts infrastructure dependencies via LLM-driven semantic analysis, persists everything to a Neo4j/Memgraph graph database, and enables cross-repo impact analysis ("blast radius") and governance policy enforcement.
Build & Run Commands
Runtime: Bun (not Node.js for development). The project uses ES modules ("type": "module").
bun install # Install dependencies
bun run build # TypeScript compilation (tsc, noEmit)
bun run dev # Run CLI directly: bun run src/cli/index.ts
bun run dev:dashboard # Dashboard dev server with live reload (http://localhost:3456)
Testing
bun run test:unit # All unit tests
bun vitest run tests/unit/path/to/file.test.ts # Single unit test
bun run test:integration # Integration tests (requires running Memgraph)
make test-eval-golden # Eval tests (LLM golden + replay-cached patterns)
make test-patterns # Pattern fixtures, deterministic subset (no LLM, no DB)
The repo organises tests by determinism level and dependency footprint. Pick the right tier when adding a feature or fixing a bug:
-
Unit tests (
tests/unit/): pure logic, no external services. Defaults for sanitizer rules, schema validation, regex guards, in-memory pipelines. Run:bun run test:unit(~7s). -
Integration tests (
tests/integration/): exercise real graph mutations against Memgraph. Required for any change tosrc/graph/mutations/, the welder (dynamic-infra-resolver.ts), or DB-backed pipelines. Run sequentially (no file parallelism). Run:bun run test:integration(~20s). -
Eval tests — agents (
tests/eval/agents/): LLM extraction quality on per-function snippets, replay-cached. Use when a fix changes prompt rules, sanitizer behaviour visible to the LLM, or the LLM output schema. Three modes viaEVAL_LLM_MODE:replay(default): cached LLM outputs fromtests/eval/.llm-cache/, ~2s, deterministiclive: real LLM calls, saves to cacherefresh: real LLM calls, overwrites cache
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 263 lines · 5,795 tokens per session scan A 6c652c6b7dc4
coderadius CLAUDE.md is an instructions file published in the GitHub repository coderadius-ai/coderadius (24 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 5,795 tokens to every session, about $0.0290 per session on Opus 5. A static security scan graded it A with 0 findings. It is 98% identical to coderadius AGENTS.md, differing in 4 lines, and is treated as a copy.
Other instructions, from other repositories
cli CLAUDE.md
Instructions for archgate/cli, covering claude.md, technology stack, commands, validation gate and git hooks (git 2.54+).
mandu AGENTS.md
Instructions for konamgil/mandu, covering mandu project guide, package manager, project structure, agent-native workflow and 배포 (release).
arkgate AGENTS.md
AGENTS.md instructions for pedroknigge/arkgate, covering arkgate enforcement (self-hosted), identity — read this first (every agent), project knowledge map, package index and surface coverage.
mandu CLAUDE.md
Instructions for konamgil/mandu, covering mandu project guide, package manager, project structure, agent-native workflow and 배포 (release).
agence AGENTS.md
AGENTS.md instructions for l-agence/agence: This repository is governed by Agence MLS policy.
agence CLAUDE.md
Claude Code instructions for l-agence/agence, covering agence: claude (anthropic) integration, why claude for agence, claude-specific configuration, model selection and environment variables.