threadknot CLAUDE.md

A project instruction file for Threadknot, a desktop app that lets several coding agents run through one interface and can show that interface in a browser on the local network. It describes the Rust backend, React frontend, agent connections, and verification requirements.

In plain words
What is it for?
Use it when developing, verifying, or renaming Threadknot, especially when working on agent connections, event handling, the shared web interface, the skills library, or stored data.
Why use it?
It gives agents the project’s architecture and compatibility rules in one place, reducing the risk of breaking data files, protocols, or the app’s rename transition.

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/smith-network-solutions/threadknot/claude-md
Clone the repo
git clone --depth 1 https://github.com/smith-network-solutions/threadknot
Per session 2,067 This file is loaded in full into every session.
When invoked 2,067 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 $0.02067 $0.02067
Opus 5 $0.01033 $0.01033
Sonnet 5 $0.00413 $0.00413
Haiku 4.5 $0.00207 $0.00207

Measured 2d ago against content hash 87a964588ba9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

threadknot 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.

CLAUDE.md · 141 lines

How it starts

The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Threadknot — every coding agent on one thread

A Tauri (Rust) desktop app that drives Claude Code, OpenAI Codex, Kimi Code, and remote hermes-agent gateways natively over their wire protocols — no terminal wrapping, no Node server — and serves the same UI to any browser on the LAN, so the app is usable from a phone.

Shipped as Armada until the 2026-08 rename. See "Rename compatibility" below before touching anything that reads from disk.

Owner: Spencer Smith (Smith Network Solutions).

Layout

  • Rust backend (src-tauri/src/): server.rs (axum :42800, token-gated /ws, serves dist/ on LAN), agents/claude.rs (claude CLI stream-json + control_request permission protocol), agents/codex.rs (codex app-server JSON-RPC), agents/kimi.rs, agents/hermes.rs (remote gateways: Runs API over HTTP/SSE), agents/mod.rs (hub: event persistence + fanout), library.rs/catalog.rs (the Library: skills written into each CLI's own skills dir + an MCP-server registry injected into every driver at spawn), store.rs (the data dir), protocol.rs (normalized event schema). Contract: docs/PROTOCOL.md.
  • Frontend: React/TS in src/, plain CSS, no state libs. Works in the Tauri webview and any phone browser (LAN URL + token in Settings popover). For the whole-screen flash — what re-rendered, what remounted, and which action caused it, readable from a headless browser — arm ?tktrace=1 and read docs/RENDER-FORENSICS.md.
  • Mobile app (mobile/): Expo SDK 57 companion — biometric-locked WebView shell with multi-server switching + push notifications. Read docs/MOBILE.md first.
  • Dispatch (dispatch.rs, exec.rs, mcp_fleet.rs): one thread hands a brief to another agent — a different harness, a different machine — and gets a report back. Read docs/USING-DISPATCH.md before changing any of it.
  • Auth: uses the installed claude / codex / kimi CLIs' own subscription logins.

Read the full file on GitHub · 141 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. 2d ago First seen · 141 lines · 2,067 tokens per session scan A 87a964588ba9

Subscribe to this mod's changes

threadknot CLAUDE.md is an instructions file published in the GitHub repository smith-network-solutions/threadknot (5 stars, last pushed 8d ago), licensed Apache-2.0. It adds 2,067 tokens to every session, about $0.0103 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.