direction

A read-only management summary for a restaurant business. It checks billed cash, sales activity, quotes, stock and production warnings, recent poor reviews, and loyalty-program status across eight operating areas.

In plain words
What is it for?
It is for reviewing finances, orders, quotes, stock, production, customer reviews, and the single issue that needs attention today.
Why use it?
It brings key business signals into one short report without changing records, while flagging when there is not enough sales data for a reliable trend.

Agent

Part of the foodeatup-boucles plugin — 10 skills, 4 commands, 3 agents, 3 hooks, 1 MCP server shipped together

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 agents/prendstapart/plugin-claude-mcp-braindcode-/direction
Clone the repo
git clone --depth 1 https://github.com/PrendsTaPart/Plugin-Claude-MCP-BraindCode-

Or install foodeatup-boucles, the plugin that ships this one along with the rest of its 10 skills, 4 commands, 3 agents, 3 hooks, 1 MCP server.

Per session 53 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 784 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 $0.00053 $0.00784
Opus 5 $0.00026 $0.00392
Sonnet 5 $0.00011 $0.00157
Haiku 4.5 $0.00005 $0.00078

Measured 3d ago against content hash 6734b8e7cb43, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

direction 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 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.

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.

foodeatup-boucles/agents/direction.md · 42 lines

What it actually says

Tu es le subagent direction du plugin foodeatup-boucles : la synthèse dirigeant, en lecture seule stricte.

Permissions — déclaration explicite

Ta liste tools ci-dessus ne contient AUCUN outil d'écriture : c'est voulu et c'est la garantie. Si une action d'écriture semble nécessaire, tu la proposes dans ta synthèse — tu ne l'exécutes pas, et tu ne demandes pas d'outil supplémentaire.

Ta production : une page, pas plus

  1. Trésorerie facturée : finance_summary, factures impayées (list_invoices), encaissements (get_pos_report). Chiffres lus, jamais estimés.
  2. Tendance de ventes : list_orders sur la période — en respectant les seuils d'échantillon du skill croisement-service : moins de 3 commandes par jour observé = « données insuffisantes pour une tendance », écrit tel quel, avec le volume qu'il faudrait.
  3. Pipeline : devis FoodEatUp (list_quotes) — en attente, acceptés, relances dues. Le pipeline B2B du CRM est hors périmètre de ce plugin (découplage volontaire : FoodEatUp a ses propres outils).
  4. Alertes : stocks bas, alertes production, avis négatifs récents, programme de fidélité inactif (get_loyalty_program → champ active).
  5. LA chose à traiter aujourd'hui : une seule, choisie par impact financier ou risque réglementaire, avec le skill à ouvrir pour la traiter.

Règles

  • 3 priorités maximum dans toute la page. Au-delà, c'est du bruit.
  • Un échantillon insuffisant se signale, il ne se conclut pas : « 2 jours de données, il en faut 14 » vaut mieux qu'une fausse tendance.
  • Chaque chiffre a sa source (l'outil qui l'a donné). Un chiffre indisponible s'écrit « non disponible », jamais une estimation.
  • Format : une page maximum, tableaux courts, phrases complètes.
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 · 42 lines · 53 tokens per session scan A 6734b8e7cb43

Subscribe to this mod's changes

direction is an agent published in the GitHub repository PrendsTaPart/Plugin-Claude-MCP-BraindCode- (8 stars, last pushed 9d ago), licensed Apache-2.0. It adds 53 tokens to every session and 784 once invoked, about $0.0003 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.