neuron-vault-write

A rule for saving durable facts about a user or project as small NeuronVault files. It classifies them as user, feedback, project, reference, or skill information.

In plain words
What is it for?
Use it when learning a lasting user preference, project decision, external-system detail, or repeatable procedure that should be remembered.
Why use it?
It preserves useful decisions, preferences, corrections, and procedures for future conversations without saving temporary chat details or sensitive information.

Cursor rule

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 rules/lezards/neuron-vault-kit/neuron-vault-write
Clone the repo
git clone --depth 1 https://github.com/lezards/neuron-vault-kit
Per session 13 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 501 The whole file, excluding the scripts and references it only reads on demand.
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 $0.00013 $0.00501
Opus 5 $0.00006 $0.00251
Sonnet 5 $0.00003 $0.00100
Haiku 4.5 $0.00001 $0.00050

Measured yesterday against content hash 62ccd3ef5643, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

neuron-vault-write 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 yesterday.

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.

integrations/cursor/neuron-vault-write.mdc · 64 lines

What it actually says

NeuronVault — escrita

Quando aprender algo que vai ser útil em conversas futuras, salve como neuron.

Quando salvar

Sinal Tipo
Usuário corrige seu approach ("não, não faz X") feedback
Usuário aprova uma escolha não-óbvia ("isso, exatamente assim") feedback
Você aprende role/stack/preferência do usuário user
Decisão de projeto, deadline, motivação não-óbvia project
Sistema externo importante (Linear, Grafana, etc) reference
Procedimento repetitivo que o usuário descreveu skill

Quando NÃO salvar

  • Padrões de código (descobre lendo o repo)
  • História do git (já está em git log)
  • Detalhes ephemeral da conversa atual
  • Algo já documentado em CLAUDE.md ou nas próprias Rules

Formato

Crie arquivo em neuron-vault/neurons/<tipo>-<slug>.md com frontmatter:

---
name: <título curto>
description: <uma linha — futuro Claude usa pra decidir relevância>
type: <user | feedback | project | reference | skill>
---

Para feedback e project, use estrutura Why/How to apply:

[a regra ou fato em uma frase]

**Why:** [motivo concreto — incidente, restrição, preferência]
**How to apply:** [quando essa regra entra em jogo]

Atualize o índice

Sempre que cria/atualiza/remove um neuron, edite neuron-vault/MEMORY.md adicionando/ajustando/removendo a linha:

- [Título](neurons/arquivo.md) — descrição em ≤150 caracteres

Não duplique

Antes de criar neuron novo, busque em MEMORY.md se já existe um sobre o mesmo tema. Se sim, atualize o existente em vez de criar duplicata.

Quando memória estiver obsoleta

Se você lê uma memória e o código contradiz, ou usuário corrige a memória, atualize ou remova — não acumule erro.

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. yesterday First seen · 64 lines · 13 tokens per session scan A 62ccd3ef5643

Subscribe to this mod's changes

neuron-vault-write is a cursor rule published in the GitHub repository lezards/neuron-vault-kit (2 stars, last pushed 15d ago), licensed MIT. It adds 13 tokens to every session and 501 once invoked, about $0.0001 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.