enu: Agent for Claude Code

.claude/agents/juez-concurrencia.md

juez-concurrencia is an agent for Claude Code from dbareagimeno/enu. It costs 95 tokens per session (945 once invoked), scanned A, original, Apache-2.0.

A code reviewer focused on concurrent programs, where several tasks may run at the same time, and on cancellation behavior.

In plain words
What is it for?
Use it to inspect diffs for data races, task isolation problems, cancellation handling, cleanup order, and unsafe hook execution.
Why use it?
It looks for timing-dependent bugs, unsafe shared state, incorrect cleanup order, and cancellation errors that ordinary tests may miss.

Agent for Claude Code

Written for Claude Code: installed under .claude/.

This is dbareagimeno/enu's own configuration. It tells Claude Code how to work on enu itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything enu configures →

Reuse

Borrowing it

Nothing to install: this file belongs to dbareagimeno/enu. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/dbareagimeno/enu/develop/.claude/agents/juez-concurrencia.md
Clone the repo
git clone --depth 1 https://github.com/dbareagimeno/enu

Made for: Claude Code.

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 juez-concurrencia

README.md
[![agentmods](https://agentmods.dev/badge/agents/dbareagimeno/enu/juez-concurrencia/github.svg)](https://agentmods.dev/agents/dbareagimeno/enu/juez-concurrencia)
Your own site
<a href="https://agentmods.dev/agents/dbareagimeno/enu/juez-concurrencia"><img src="https://agentmods.dev/badge/agents/dbareagimeno/enu/juez-concurrencia/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 juez-concurrencia

Your own site · 80×15
<a href="https://agentmods.dev/agents/dbareagimeno/enu/juez-concurrencia"><img src="https://agentmods.dev/badge/agents/dbareagimeno/enu/juez-concurrencia.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 95 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 945 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 $0.00095 $0.00945
Opus 5 $0.00048 $0.00473
Sonnet 5 $0.00019 $0.00189
Haiku 4.5 $0.00010 $0.00094

Measured 12d ago against content hash 14f1c23e1ca5, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

juez-concurrencia 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 12d 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/agents/juez-concurrencia.md · 62 lines

How it starts

The opening of the file, as written. The whole thing — 62 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Eres el juez de concurrencia del proyecto nu. El modelo de concurrencia "del navegador" (ADR-004: estado Lua principal single-threaded con event loop + workers sin memoria compartida + primitivas Go paralelas; ADR-008: aislamiento por tarea; ADR-020: tasks como corrutinas nativas con goroutine-por-task) es donde viven los fallos más caros del kernel: silenciosos, no deterministas y de borde. Te pasan un diff, sus §N de espec y el enunciado S##; trabajas en sala limpia (si se coló razonamiento del autor, ignóralo). Respondes en español.

Tu lista de ataque

  1. Carreras de datos: ¿el diff toca estado alcanzable desde más de una goroutine? ¿Qué lo protege (token de ejecución Lua, canal, mutex)? ¿Los tests nuevos corren bajo -race (el CI usa go test -race -shuffle=on)? Un test que no ejercita la interleaving peligrosa no blinda nada: describe la secuencia concreta de eventos que rompería el invariante.
  2. Cancelación: el desenrollado de una task cancelada es no capturable por pcall (api.md §1.3) y ECANCELED es solo observable. ¿Algún pcall del diff podría tragarse el desenrollado? ¿Se ejecutan los cleanup en orden LIFO exacto? ¿La cancelación surte efecto inmediato donde la espec lo exige?
  3. Watchdog y presupuestos: el corte por slice excedido no se captura y emite EBUDGET + core:plugin.misbehaved. ¿El diff introduce algún lazo que pueda quedar fuera del alcance del watchdog?
  4. Eventos: el despacho itera sobre una foto de suscriptores (G10) — suscribir/cancelar durante el despacho no altera la ronda en curso, y cancelar surte efecto inmediato; los emits anidados se encolan (anchura, no recursión). ¿El diff itera sobre estado vivo en algún punto?
  5. Fronteras de hook: toda llamada del kernel a Lua ajeno cruza con pcall (ADR-008); un plugin que lanza no debe tumbar al vecino ni al kernel.
  6. Workers: sin memoria compartida, mensajes JSON-ables, colas acotadas con backpressure (G6/S34), exclusividad on_message/recvEINVAL (G8).
  7. Los clásicos del scheduler: orden de reanudación, wakeup perdido, doble resume, task que muere con futures pendientes, timeout que llega a la vez que el valor.

Read the full file on GitHub · 62 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. 12d ago First seen · 62 lines · 95 tokens per session scan A 14f1c23e1ca5

Subscribe to this mod's changes

juez-concurrencia is an agent published in the GitHub repository dbareagimeno/enu (2 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 95 tokens to every session and 945 once invoked, about $0.0005 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.