vscode chat.instructions.md

Guidelines for adding chat and AI features to VS Code, including when those features must be hidden if AI is disabled.

In plain words
What is it for?
Use it when adding or reviewing chat surfaces, AI commands, menus, views, or other AI-related interface elements.
Why use it?
It prevents users from seeing AI-related commands, views, or controls after they have disabled AI features or indicated they do not want them.

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/microsoft/vscode/chat
Clone the repo
git clone --depth 1 https://github.com/microsoft/vscode

Made for: GitHub Copilot.

Per session 370 This file is loaded in full into every session.
When invoked 370 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 $0.00370 $0.00370
Opus 5 $0.00185 $0.00185
Sonnet 5 $0.00074 $0.00074
Haiku 4.5 $0.00037 $0.00037

Measured yesterday against content hash 2f7ddfb21166, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

vscode chat.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/instructions/chat.instructions.md · 27 lines

What it actually says

  • When adding a new chat/AI feature like a new surface where chat or agents appear, a new AI command, etc, these features must not show up for users when they've disabled AI features. The best way to do this is to gate the feature on the context key ChatContextKeys.enabled via a when clause.
  • When doing a code review for code that adds an AI feature, please ensure that the feature is properly gated.

Hiding AI Features When Disabled

When surfacing a UI AI feature, ensure the feature hides when chat.disableAIFeatures is set:

  • UI Hiding: Use ChatContextKeys.enabled in when conditions to conditionally show/hide UI elements (commands, views, menu items, etc.)
    • Example: when: ChatContextKeys.enabled in action/command registration
    • This context key is false when AI features are disabled
  • Programmatic Hiding: Check IChatEntitlementService.sentiment.hidden to determine if AI features should be hidden
    • sentiment.hidden is true when the user signals no intent in using Chat
    • This should not only disable Chat but also hide all of its UI

This ensures consistency when implementing AI-powered UI functionality across the codebase.

Rendering Agent-Provided Markdown

  • Determine content provenance before rendering markdown. Content authored by an agent must be treated as untrusted; localized VS Code text, tool implementation messages, extension metadata, and other internally authored content do not automatically require the same restrictions.
  • Render agent-provided markdown with ChatContentMarkdownRenderer. When a generic markdown renderer is required, apply getChatMarkdownRenderOptions().
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 · 27 lines · 370 tokens per session scan A 2f7ddfb21166

Subscribe to this mod's changes

vscode chat.instructions.md is an instructions file published in the GitHub repository microsoft/vscode (190,061 stars, last pushed yesterday), licensed MIT. It adds 370 tokens to every session, about $0.0019 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 2026-08-30.