roboot CLAUDE.md

Repository instructions for Roboot, a macOS AI agent hub that manages Claude Code sessions and provides a web and voice interface.

In plain words
What is it for?
Use them when developing or maintaining Roboot, including its web server, session controls, remote access, and voice features.
Why use it?
They explain the project's architecture, commands, and main console so an agent can change the code with the system's structure in mind.

Instructions file

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/tyxben/roboot/claude-md
Clone the repo
git clone --depth 1 https://github.com/tyxben/roboot
Per session 8,216 This file is loaded in full into every session.
When invoked 8,216 The same file — it is already loaded in full.
Security scan F 4 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 $0.08216 $0.08216
Opus 5 $0.04108 $0.04108
Sonnet 5 $0.01643 $0.01643
Haiku 4.5 $0.00822 $0.00822

Measured 3d ago against content hash 9b365e6827c0, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade F, and why

roboot CLAUDE.md scanned grade F with 4 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 3d 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.

Downloads and executes remote codehighSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

Residual: Arcana's MCP layer classifies a tool's side-effect by a name/description keyword heuristic (`_infer_side_effect`, default READ), so a write tool whose name dodges the keywords classifies READ and never reaches

Encoded or obfuscated payloadhighSupply chain

base64 or hex that is decoded and executed hides what actually runs from anyone reading the file.

Residual: Arcana's MCP layer classifies a tool's side-effect by a name/description keyword heuristic (`_infer_side_effect`, default READ), so a write tool whose name dodges the keywords classifies READ and never reaches

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

The danger detector applies NFKC + ANSI strip + null-byte normalization before matching, with a 16 KB hard cap on detector input (ReDoS guard). Allowlist at `~/.roboot/tool_allowlist.json` (per-machine, gitignored) does

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

Residual: Arcana's MCP layer classifies a tool's side-effect by a name/description keyword heuristic (`_infer_side_effect`, default READ), so a write tool whose name dodges the keywords classifies READ and never reaches
CLAUDE.md · 301 lines

How it starts

The opening of the file, as written. The whole thing — 301 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 is Roboot

A personal AI agent hub that runs on macOS. It manages multiple Claude Code sessions in iTerm2, executes shell commands, and provides a JARVIS-like voice interface -- all powered by the Arcana agent framework. Supports remote access via LAN, Telegram, and a Cloudflare Worker relay.

Architecture

server.py (FastAPI)              <- Main entry point
+-- WebSocket /ws                <- Streaming chat (LLM_CHUNK events)
+-- REST /api/sessions/*         <- Direct iTerm2 session control
+-- REST /api/tts                <- Edge TTS (text -> mp3)
+-- REST /api/relay-info         <- Relay pairing URL + expiry
+-- REST /api/relay-refresh      <- Rotate relay token (POST)
+-- REST /api/relay-revoke       <- Kick all clients + rotate (POST)
+-- REST /api/relay-qr           <- QR code PNG for relay pairing URL
+-- REST /api/network-info       <- Local IP addresses
+-- REST /api/qr-code            <- QR code PNG for LAN URL
+-- REST /api/filevault-status   <- macOS FileVault probe (fdesetup)
+-- REST /api/chat-history-wipe  <- Drop all chat rows + VACUUM (POST)
+-- Static /static/console.html  <- Unified web console

relay_client.py                  <- Relay WebSocket client (connects to CF Worker)
iterm_bridge.py                  <- Persistent iTerm2 Python API connection
network_utils.py                 <- IP detection, QR code generation
soul.md                          <- Assistant's self-modifiable identity
config.yaml                      <- API keys + provider config (gitignored)

text_utils.py                    <- Shared helpers (extract_spoken_text, ...)
tts_synth.py                     <- Edge TTS helper shared by /api/tts + mobile relay
soul_review.py                   <- Review gate for soul.md overwrites (off/log/confirm)
tool_guard.py                    <- Approval gate for agent tool calls (off/log/confirm) — gates run_command via Arcana confirmation_callback
mcp_bootstrap.py                 <- Connect external MCP servers at startup; register their tools (gated as external writes)
proactive.py                     <- Proactive daily briefing (read-only autonomous agent turn on a daily clock)
filevault_status.py              <- macOS fdesetup probe for the console warning banner

tools/                           <- Arcana tools (agent's capabilities)
+-- shell.py                     <- Terminal command execution
+-- claude_code.py               <- iTerm2 session list/read/send/create
+-- vision.py                    <- Camera capture + screenshot + face recognition
+-- face_db.py                   <- Face encoding storage and matching (.faces/)
+-- soul.py                      <- Self-modification + user memory
+-- voice_switch.py              <- Agent tool: change Telegram TTS voice
+-- scheduler.py                 <- Delayed/recurring reminders (.reminders.db, per-origin dispatcher)
+-- files.py                     <- read_file/write_file/edit_file (own path deny-policy; writes gated)
+-- web.py                       <- web_fetch/web_search (SSRF-guarded, keyless DuckDuckGo)

adapters/                        <- I/O adapters
+-- telegram_bot.py              <- Remote control via Telegram (voice I/O)
+-- voice.py                     <- Local mic STT + macOS TTS (CLI --voice)
+-- voice_prefs.py               <- Per-Telegram-user TTS voice store
+-- tts_streamer.py              <- Edge TTS -> parallel OGG/Opus synthesis
+-- keyboard.py                  <- Terminal text input
+-- stt/                         <- Pluggable speech-to-text backends
    +-- mlx.py                   <- mlx-whisper (Apple Silicon, default)
    +-- google.py                <- speech_recognition -> Google Web Speech
    +-- noop.py                  <- backend: none

scripts/
+-- setup.sh                     <- One-command installer (auto-uses uv)

relay/                           <- Cloudflare Worker relay server
+-- src/index.ts                 <- Worker entry point, routing, rate limiting
+-- src/relay-session.ts         <- Durable Object: daemon<->client session mgmt
+-- src/pair-page.ts             <- HTML pairing page served to mobile clients
+-- wrangler.toml                <- Cloudflare deployment config

Read the full file on GitHub · 301 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. 3d ago First seen · 301 lines · 8,216 tokens per session scan F 9b365e6827c0

Subscribe to this mod's changes

roboot CLAUDE.md is an instructions file published in the GitHub repository tyxben/roboot (81 stars, last pushed 2mo ago), licensed MIT. It adds 8,216 tokens to every session, about $0.0411 per session on Opus 5. A static security scan graded it F with 4 findings (downloads and executes remote code, encoded or obfuscated payload, recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other instructions, from other repositories

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,345 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

next.js AGENTS.md

Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens