neurarch-skill

neurarch-skill is a skill for Claude Code, Codex from neurarch-ai/neurarch-mcp. It costs 125 tokens per session (2,174 once invoked), scanned A, original, MIT.

A controlled workflow for inspecting and changing a neural-network design stored as a Neurarch graph file. A neural network is a machine-learning model made of connected processing layers.

In plain words
What is it for?
It helps summarize a model, audit or debug its structure, find parameter and computation hotspots, trace dependencies, and extend or refactor the graph. Changes are made through the Neurarch MCP tools rather than by directly editing the file.
Why use it?
It keeps the graph as the single source of truth and requires evidence from the connected tools before architecture changes. This reduces the risk of edits that conflict with the live model or its conventions.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit It helps summarize a model, audit or debug its structure, find parameter and computation hotspots, trace dependencies, and extend or refactor the graph. Changes are made through the Neurarch MCP tools rather than by directly editing the file.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/neurarch-ai/neurarch-mcp/neurarch-skill
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.

Any agent
npx skills add neurarch-ai/neurarch-mcp --skill neurarch-skill
Clone the repo
git clone --depth 1 https://github.com/neurarch-ai/neurarch-mcp

Made for: Claude Code, Codex.

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 neurarch-skill

README.md
[![agentmods](https://agentmods.dev/badge/skills/neurarch-ai/neurarch-mcp/neurarch-skill/github.svg)](https://agentmods.dev/skills/neurarch-ai/neurarch-mcp/neurarch-skill)
Your own site
<a href="https://agentmods.dev/skills/neurarch-ai/neurarch-mcp/neurarch-skill"><img src="https://agentmods.dev/badge/skills/neurarch-ai/neurarch-mcp/neurarch-skill/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for neurarch-skill

Your own site · 80×15
<a href="https://agentmods.dev/skills/neurarch-ai/neurarch-mcp/neurarch-skill"><img src="https://agentmods.dev/badge/skills/neurarch-ai/neurarch-mcp/neurarch-skill.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 125 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,174 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 $0.00125 $0.02174
Opus 5 $0.00063 $0.01087
Sonnet 5 $0.00025 $0.00435
Haiku 4.5 $0.00013 $0.00217

Measured 8d ago against content hash 95ee026809d0, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

neurarch-skill 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 8d 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.

skills/neurarch-skill/SKILL.md · 130 lines

How it starts

The opening of the file, as written. The whole thing — 130 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Neurarch Graph Loop

Use the Neurarch model graph as the single source of truth for a network's structure. Move every architecture question or edit through one bounded, evidence-gated pass using only the neurarch-mcp tools exposed by the host. Numbers come from tools, not from the model's memory of similar architectures.

Read the references

Operating contract

  1. Use only the connected neurarch-mcp server for graph reads and writes. Do not hand-edit the .neurarch.json file while the server is connected: the server holds an in-memory copy, and direct file edits either get overwritten by save_model or silently diverge from what the tools report.
  2. Treat the connected tool schemas as authoritative. MCP hosts may namespace tool names; match the terminal tool name and semantics rather than assuming a prefix.
  3. Discover the model's vocabulary at runtime: layer types, layer names, scopes, blocks, and hyperparameters come from the live graph, never from another model, a template, or memory of a similar architecture.
  4. Frame one authorized edit at a time, including the evidence that will prove it landed and the condition that would stop the loop.
  5. Make the smallest mutation that can complete that edit, then re-read the affected part of the graph.
  6. Treat a write tool's success response as a claim. Only a follow-up read (get_layer, list_connections, validate_model) confirms the intended state.
  7. Run validate_model before the first edit and record the result. Pre-existing issues belong to the baseline; only new issues belong to your change.
  8. Call save_model after verified mutations. In-memory edits are lost on server restart, and with --watch an external save from the Neurarch app overwrites them.
  9. Preserve the model's existing naming style, scope structure, and design notes unless the user explicitly asks to change them.

Read the full file on GitHub · 130 lines

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 8d ago First seen · 130 lines · 125 tokens per session scan A 95ee026809d0

Subscribe to this mod's changes

neurarch-skill is a skill published in the GitHub repository neurarch-ai/neurarch-mcp (1 stars, last pushed 2d ago), licensed MIT. It adds 125 tokens to every session and 2,174 once invoked, about $0.0006 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 2026-09-04.