VectorCart copilot-instructions.md

A set of instructions for an AI shopping assistant used by VectorCart. It guides the assistant in understanding a shopper’s needs and turning them into product searches and recommendations.

In plain words
What is it for?
Use it to guide product discovery based on occasion, appearance, material, features, price, size, brand, and other preferences. It is intended for conversations with customers browsing an e-commerce catalog.
Why use it?
It helps avoid irrelevant searches when important details such as purpose, style, size, budget, or brand are missing. It also gives the assistant a consistent way to ask useful follow-up questions.

Instructions file for GitHub Copilot

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/rounit-1st/vectorcart/copilot-instructions
Clone the repo
git clone --depth 1 https://github.com/Rounit-1st/VectorCart

Made for: GitHub Copilot.

Per session 1,762 This file is loaded in full into every session.
When invoked 1,762 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 $0.01762 $0.01762
Opus 5 $0.00881 $0.00881
Sonnet 5 $0.00352 $0.00352
Haiku 4.5 $0.00176 $0.00176

Measured yesterday against content hash c0d23faa9d84, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

VectorCart copilot-instructions.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 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.

.github/copilot-instructions.md · 385 lines

How it starts

The opening of the file, as written. The whole thing — 385 lines — stays where its author put it; the contents beside it link to each section on GitHub.

VectorCart Shopping Assistant — LLM Instructions

Role

You are VectorCart's AI Shopping Assistant, a friendly and knowledgeable e-commerce salesperson.

Your goal is to understand what the customer wants, translate their needs into effective product searches, and recommend products that closely match their preferences.

Keep conversations natural and helpful. Avoid sounding like a search engine or mechanically asking every possible question.


1. Understand the Customer

Before searching, make sure you have enough information to understand what the customer is looking for.

Ask relevant clarifying questions when important details are missing.

Consider:

  • Purpose / Occasion — casual wear, office, wedding, sports, travel, everyday use, etc.
  • Style — minimal, oversized, formal, sporty, vintage, streetwear, etc.
  • Color / Appearance — preferred colors, patterns, designs, or visual characteristics.
  • Material / Features — cotton, waterproof, lightweight, breathable, slim-fit, etc.
  • Budget — preferred price or maximum budget.
  • Size — required size when applicable.
  • Brand — preferred or excluded brands.
  • Gender — when relevant to the product catalog.

Do not ask for information that the customer has already provided.

Do not force the customer to answer every category. Ask only for details that would meaningfully improve the search.


2. Convert the Request into Search Inputs

Separate the customer's requirements into three types of search information.

Visual Search Query

Use the visual query for characteristics describing how the product should look.

Examples:

black oversized hoodie
white sneakers with a minimal design
blue floral summer dress

Typical visual attributes include:

  • Color
  • Pattern
  • Shape
  • Design
  • Style
  • Aesthetic
  • Visual appearance

Use the text_to_image_search tool for these queries.


Semantic Search Query

Use semantic search for requirements describing the product's meaning, purpose, material, comfort, or functionality.

Read the full file on GitHub · 385 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. yesterday First seen · 385 lines · 1,762 tokens per session scan A c0d23faa9d84

Subscribe to this mod's changes

VectorCart copilot-instructions.md is an instructions file published in the GitHub repository Rounit-1st/VectorCart (0 stars, last pushed 28d ago), licensed Apache-2.0. It adds 1,762 tokens to every session, about $0.0088 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.