Switchboard CLAUDE.md

Switchboard CLAUDE.md is an instructions file for coding agents from johnjanthony/Switchboard. It costs 7,188 tokens per session, scanned D, original, MIT.

Project instructions for agents working on Switchboard, a local hub that tracks coding-agent sessions, sends updates to a phone, and lets phone commands control agents. The document explains the project structure, local setup, and testing.

In plain words
What is it for?
Use it when asking an agent to modify, run, or test Switchboard, including its local server, dashboard, phone connection, and agent-control features.
Why use it?
An agent joining the project can learn its architecture and working conventions from one place instead of guessing how the parts fit together.

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/johnjanthony/switchboard/claude-md
Clone the repo
git clone --depth 1 https://github.com/johnjanthony/Switchboard

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 Switchboard CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/johnjanthony/switchboard/claude-md.svg)](https://agentmods.dev/instructions/johnjanthony/switchboard/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/johnjanthony/switchboard/claude-md"><img src="https://agentmods.dev/badge/instructions/johnjanthony/switchboard/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 7,188 This file is loaded in full into every session.
When invoked 7,188 The same file — it is already loaded in full.
Security scan D 3 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 $0.07188 $0.07188
Opus 5 $0.03594 $0.03594
Sonnet 5 $0.01438 $0.01438
Haiku 4.5 $0.00719 $0.00719

Measured 3d ago against content hash ee57643c4555, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade D, and why

Switchboard CLAUDE.md scanned grade D with 3 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.

**Diagnostic:** `curl -s http://localhost:9876/healthz | python -m json.tool` reports listener supervision state, dispatch-loop crash counts, and pending-question state.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

Antigravity (agy) sessions wire four hook events - PreInvocation, PreToolUse, PostToolUse, Stop - via the chezmoi-managed `~/.gemini/config/hooks.json`, not this plugin. There are no SessionStart/SessionEnd equivalents:

Makes network callslowCapability

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

**Diagnostic:** `curl -s http://localhost:9876/healthz | python -m json.tool` reports listener supervision state, dispatch-loop crash counts, and pending-question state.
CLAUDE.md · 316 lines

How it starts

The opening of the file, as written. The whole thing — 316 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Switchboard — Agent Orientation

Switchboard is the mission-control hub for a multi-agent workstation: it tracks Claude Code sessions and conversations, routes questions and notifications to a phone, mirrors telemetry to ambient surfaces (Operator, Watchtower), and dispatches phone-issued commands (spawn, combine, away mode). Its founding feature - and still the core protocol - is away mode: agents pause mid-task and request human input, answered from the phone.

This file orients any agent (Claude Code, etc.) working on Switchboard itself or consuming its tools. Switchboard ships as a Claude Code plugin; see Setup below.

Design history lives under docs/; the current end-to-end design is docs/switchboard-design-spec-comprehensive.md, and the dated specs under docs/superpowers/specs/ record how each subsystem evolved. The newest dated spec wins where two overlap.

Project shape

Single Python process, one asyncio event loop, MCP HTTP server on localhost:9876, with a Firebase backend (Android + Realtime Database). The server also serves Switchboard Operator, a launch-on-demand web cockpit (a zero-build Preact+htm app in dashboard/), at /dashboard, plus a small widget-facing GET /stats roll-up. The dashboard talks to Firebase RTDB directly and reads /healthz for its health panel.

Away mode is the founding feature, not the whole product: ask/notify blocking semantics are away-mode-scoped (at-desk interaction uses the terminal), while session tracking, telemetry fan-out, Operator, and Watchtower are always-on ambient surfaces.

The Registry is in-memory. The pending-request index is keyed by (conversation_id, cli_session_id) tuples where conversation_id is a conv-<uuid> string from Registry.conversations — not a filesystem path; answers resolve by (conversation_id, request_id). Pending ask_human futures die on restart, but the questions survive: hydration rebuilds pending_questions records as parked (future-less) pendings, an arriving answer resolves them with a history write plus a session notice, and unanswered ones expire at the 72h retention horizon (chunk 7). Conversations (the persistence unit) survive restart via Firebase hydration — see server/hydration.py.

Read the full file on GitHub · 316 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 · 316 lines · 7,188 tokens per session scan D ee57643c4555

Subscribe to this mod's changes

Switchboard CLAUDE.md is an instructions file published in the GitHub repository johnjanthony/Switchboard (0 stars, last pushed 9d ago), licensed MIT. It adds 7,188 tokens to every session, about $0.0359 per session on Opus 5. A static security scan graded it D with 3 findings (downloads and executes remote code, reads agent configuration directories, makes network calls). 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

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