arandu CLAUDE.md

arandu CLAUDE.md is an instructions file for Claude Code from vdittgen/arandu. It costs 2,103 tokens per session, scanned A, original, Apache-2.0.

Project instructions for Arandu, a privacy-focused desktop AI system that keeps AI processing on the user's device with local Ollama models.

In plain words
What is it for?
Use them when developing Arandu's Rust, Python, or TypeScript components.
Why use it?
They explain the project's architecture, privacy boundaries, data systems, and coding conventions before changes are made.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md.

Install

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.

agentmods
npx agentmods add instructions/vdittgen/arandu/claude-md
Clone the repo
git clone --depth 1 https://github.com/vdittgen/arandu

Made for: Claude Code.

Wrote 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.

agentmods badge for arandu CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/vdittgen/arandu/claude-md.svg)](https://agentmods.dev/instructions/vdittgen/arandu/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/vdittgen/arandu/claude-md"><img src="https://agentmods.dev/badge/instructions/vdittgen/arandu/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 2,103 This file is loaded in full into every session.
When invoked 2,103 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.02103 $0.02103
Opus 5 $0.01052 $0.01052
Sonnet 5 $0.00421 $0.00421
Haiku 4.5 $0.00210 $0.00210

Measured 5d ago against content hash c510b79d5b62, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

arandu 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 5d 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.

CLAUDE.md · 139 lines

How it starts

The opening of the file, as written. The whole thing — 139 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Arandu

Privacy-first AI operating system that runs as a native desktop app. Open source (Apache-2.0), local inference only (Ollama).

Architecture

  • App shell: Tauri (Rust backend + web frontend)
  • Analytical DB: SQLite (embedded, WAL mode)
  • Graph DB: Kuzu (embedded, relationships/entities)
  • Vector search: ChromaDB (embedded, semantic similarity)
  • Pipeline: Manifest-driven Python (staging → intermediate → marts)
  • LLM: Local Ollama (all inference stays on device). Routing policy lives in src/agents/firewall/egress_firewall.py; see docs/PRIVACY.md. The provider factory in src/models/llm_provider.py builds the local Ollama provider — there is no remote/cloud provider in this codebase.
  • Frontend: React + Tailwind inside Tauri webview
  • Languages: Rust (firewall, IPC), Python (pipeline, ML, agents), TypeScript (UI)

Top-level packages

Package Bounded context
src/core/ Storage (sqlite/kuzu/chromadb), query engine, data layer, CLI dispatcher
src/agents/ Brain Agent, fact learner, insight generator, proactive intelligence, message triage/eval
src/agent_runtime/ Sandboxed agent runner, context, sensitivity guard, skills (3rd-party isolation)
src/extensions/mcp/ MCP protocol: client, installer, tool classifier
src/extensions/connectors/ Connector lifecycle: catalog, registry, manager, scheduler, requirements
src/extensions/bridges/ Native bridges: apple/, whatsapp/
src/extensions/ingestion/ Sync engine, adapters, transforms, schema discovery, model generator
src/extensions/builtin/ Built-in agents (weekly_digest, relationship_tracker)
src/notifications/ Orchestrator, notifier, reply handler, preferences
src/pipeline/ Manifest, runner, executor, brain, auditor, worker
src/models/ LLM provider abstraction, voice, sensitivity classifier
src/interface/ React frontend (pages, components, hooks)
src-tauri/src/firewall/ Audit chain (Rust reader of ~/.arandu/data/audit.jsonl)

Read the full file on GitHub · 139 lines

Changes

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.

  1. 5d ago First seen · 139 lines · 2,103 tokens per session scan A c510b79d5b62

Subscribe to this mod's changes

arandu CLAUDE.md is an instructions file published in the GitHub repository vdittgen/arandu (2 stars, last pushed 27d ago), licensed Apache-2.0. It adds 2,103 tokens to every session, about $0.0105 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 2026-08-31.

Related

Other instructions, from other repositories

fak AGENTS.md

AGENTS.md instructions for anthony-chaudhary/fak, covering agents.md — orientation for coding agents, default scope when launched from this repo, delegate real work; keep the coordinator context clean, scope discipline for smaller models: subdivide or abstain and native inference performance invariant.

anthony-chaudhary/fak · 7,871 tokens

moosestack copilot-instructions.md

Instructions for 514-labs/moosestack, covering moose repository guide, code standards, required before each commit, development flow and repository structure.

514-labs/moosestack · 519 tokens

fak CLAUDE.md

Claude Code instructions for anthony-chaudhary/fak, covering claude.md, the canonical instructions, the five hard rules and doc map.

anthony-chaudhary/fak · 970 tokens

fak GEMINI.md

Gemini CLI instructions for anthony-chaudhary/fak: The canonical agent instructions for this repo are in AGENTS.md — read it first for build/test/run, the repo map, and the rules. Curated doc map: llms.txt. Contributor contract: CONTRIBUTING.md.

anthony-chaudhary/fak · 904 tokens

fak copilot-instructions.md

Copilot instructions for anthony-chaudhary/fak: The canonical agent instructions for this repo are in AGENTS.md — read it first for build/test/run, the repo map, and the rules. Curated doc map: llms.txt. Contributor contract: CONTRIBUTING.md.

anthony-chaudhary/fak · 503 tokens

moosestack AGENTS.md

Instructions for 514-labs/moosestack, covering agents.md, build & development commands, all languages, rust and typescript.

514-labs/moosestack · 1,794 tokens