VTCode copilot-instructions.md

Repository instructions for VT Code, a Rust terminal coding agent that supports several AI providers and programming languages. They describe its workspace, main modules, build commands, tests, and code conventions.

In plain words
What is it for?
Use them when modifying VT Code, running its tests or checks, formatting Rust code, or working on its model, tool, configuration, or parser modules.
Why use it?
They show an agent how to build and test the Rust workspace and where features such as language parsing, terminal execution, model connections, and tool support are organized.

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/vinhnx/vtcode/copilot-instructions
Clone the repo
git clone --depth 1 https://github.com/vinhnx/VTCode

Made for: GitHub Copilot.

Per session 2,722 This file is loaded in full into every session.
When invoked 2,722 The same file — it is already loaded in full.
Security scan C 1 finding. 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 $0.02722 $0.02722
Opus 5 $0.01361 $0.01361
Sonnet 5 $0.00544 $0.00544
Haiku 4.5 $0.00272 $0.00272

Measured yesterday against content hash 0df4969dd082, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

VTCode copilot-instructions.md scanned grade C with 1 finding 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.

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

- For operations that are potentially destructive (e.g., `git reset --hard`, `git push --force`, `rm -rf`), require explicit confirmation: supply `confirm=true` in the tool input or include an explicit `--confirm` flag.
.github/copilot-instructions.md · 263 lines

How it starts

The opening of the file, as written. The whole thing — 263 lines — stays where its author put it; the contents beside it link to each section on GitHub.

VT Code Agent Guidelines

VT Code: Rust terminal coding agent with modular architecture, multi-LLM support (OpenAI, Anthropic, Gemini), tree-sitter parsing for 6+ languages.

Build & Test Commands

# Preferred (faster with cargo-nextest)
cargo nextest run           # Run all tests (3-5x faster)
cargo nextest run -p vtcode-core  # Single package
cargo t                     # Alias for nextest run
cargo tq                    # Quick profile (no retries)

# Standard cargo commands
cargo check                 # Fast compile check
cargo clippy                # Lint (strict Clippy rules)
cargo fmt                   # Format code

# Fallback (if nextest unavailable)
cargo ts                    # Alias for standard cargo test

Architecture & Key Modules

  • Workspace: crates/codegen/vtcode-core/ (library) + src/main.rs (binary) + 9 workspace crates
  • Core: llm/ (multi-provider), tools/ (trait-based), config/ (TOML-based)
  • Integrations: Tree-sitter, PTY execution, ACP/MCP protocol, Gemini/OpenAI/Anthropic APIs

Code Style & Conventions

  • Naming: snake_case functions/vars, PascalCase types (standard Rust)
  • Error Handling: anyhow::Result<T> + anyhow::Context; NO unwrap()
  • Constants: Use crates/codegen/vtcode-core/src/config/constants.rs (never hardcode, especially model IDs)
  • Config: Read from vtcode.toml at runtime
  • Docs: Markdown ONLY in ./docs/; use docs/models.json for latest LLM models
  • Formatting: 4-space indentation, early returns, simple variable names
  • Ownership first: Prefer owned values and borrows; reach for Rc<T> / Arc<T> only when multiple owners genuinely need to keep data alive
  • Shared ownership: Rc<T> is single-threaded; Arc<T> is for cross-thread/task sharing. Prefer immutable sharing before RefCell, Mutex, or RwLock
  • Cycles: Use Weak<T> / Arc::downgrade() for back-references or task-parent links so shared graphs can drop cleanly

See Also

Read the full file on GitHub · 263 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 · 263 lines · 2,722 tokens per session scan C 0df4969dd082

Subscribe to this mod's changes

VTCode copilot-instructions.md is an instructions file published in the GitHub repository vinhnx/VTCode (825 stars, last pushed 2d ago), licensed Apache-2.0. It adds 2,722 tokens to every session, about $0.0136 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.