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/nexus-router/nexus/claude-mdgit clone --depth 1 https://github.com/Nexus-Router/nexusWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/nexus-router/nexus/claude-md)<a href="https://agentmods.dev/instructions/nexus-router/nexus/claude-md"><img src="https://agentmods.dev/badge/instructions/nexus-router/nexus/claude-md.svg" alt="Measured on agentmods" height="20"></a>What 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 5.1 | $0.05592 | $0.05592 |
| Opus 5 | $0.02796 | $0.02796 |
| Sonnet 5 | $0.01118 | $0.01118 |
| Haiku 4.5 | $0.00559 | $0.00559 |
Grade A, and why
nexus 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 6d 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.
How it starts
The opening of the file, as written. The whole thing — 694 lines — stays where its author put it; the contents beside it link to each section on GitHub.
LLM Code Editing Guidelines for Nexus
When editing code in this repository, you are working on Nexus, an AI router that aggregates MCP (Model Context Protocol) servers and LLMs. This system helps reduce tool proliferation by intelligently routing and indexing available tools.
Commit Messages
- Follow the Conventional Commits 1.0.0 specification for every commit message.
- Use recent commits (
git log) as a reference for tone and formatting if you need examples. - Structure messages as
type(issue number): subject, where - Common
typevalues includefeat,fix,chore,docs,refactor,test, andbuild; pick the one that best matches the change. - The issue number is a linear issue number, in the format of
gb-1234. The user should give you this in the prompt, but if you don't have it, you can ask the user for it. - Keep the subject in the imperative mood, ≤72 characters, and avoid punctuation at the end.
Domain Context
- MCP (Model Context Protocol): A protocol for connecting AI models with external tools and data sources
- AI Router: Nexus acts as an intelligent intermediary between LLMs and multiple MCP servers
- Tool Indexing: Uses Tantivy (full-text search engine) to create searchable indexes of available tools
- Dynamic vs Static Tools: Static tools are shared across users; dynamic tools require user authentication
Key Technologies
- Rust: The primary programming language
- Axum: Web framework for HTTP routing and middleware
- Tantivy: Full-text search engine for tool indexing
- Tokio: Async runtime
- Serde: Serialization/deserialization for JSON/TOML
- RMCP: Rust MCP client/server implementation
- Anyhow: Error handling with context and backtrace
- Reqwest: HTTP client for external API calls
- JWT-Compact: JWT token handling for authentication
- Tower/Tower-HTTP: Middleware and service layers
- Rustls: TLS implementation for secure connections
- Insta: Snapshot testing framework
- Clap: Command-line argument parsing
- Logforth: Structured logging with tracing support
- Docker Compose: For integration testing with Hydra OAuth2 server
- Governor: Rate limiting with token bucket algorithm
- Mini-moka: In-memory caching for rate limit buckets
- Redis: Redis support for distributed rate limiting
- Deadpool: Connection pooling
- OpenTelemetry: Observability with metrics collection and OTLP export
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.
- 6d ago First seen · 694 lines · 5,592 tokens per session scan A d6c1758d75cb
nexus CLAUDE.md is an instructions file published in the GitHub repository Nexus-Router/nexus (435 stars, last pushed 5mo ago), licensed MPL-2.0. It adds 5,592 tokens to every session, about $0.0280 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
openinference AGENTS.md
AGENTS.md instructions for Arize-ai/openinference, covering claude.md, repository overview, repository structure, essential commands by language and python.
atomic-agents AGENTS.md
Instructions for Eigenwise/atomic-agents, covering atomic agents monorepo, what is this repository?, core philosophy, monorepo structure and package details.
synalinks AGENTS.md
AGENTS.md instructions for SynaLinks/synalinks, covering agents.md, project overview, development commands, install dependencies and run tests with coverage.
gateway CLAUDE.md
Instructions for Portkey-AI/gateway, covering claude.md, project overview, development commands, core development and testing.
langfuse AGENTS.md
AGENTS.md instructions for langfuse/langfuse, a project described as: 🪢 Open source AI engineering platform: LLM evals, observability, metrics, prompt management, playground, datasets. Integrates with OpenTelemetry, LangChain, OpenAI SDK, LiteLLM, and more. 🍊YC W23.
langfuse CLAUDE.md
Claude Code instructions for langfuse/langfuse, a project described as: 🪢 Open source AI engineering platform: LLM evals, observability, metrics, prompt management, playground, datasets. Integrates with OpenTelemetry, LangChain, OpenAI SDK, LiteLLM, and more. 🍊YC W23.