agentmods.dev

corvus: Instructions file for Claude Code

CLAUDE.md

corvus CLAUDE.md is an instructions file for Claude Code from Ambar-Gupta22/corvus. It costs 5,806 tokens per session, scanned C, original, MIT.

Repository guidance for Corvus, an offline-capable C++17 runtime for building AI agents inside applications such as robots, games, drones, and trading systems. It explains the project's goals, design priorities, and reference documents.

In plain words
What is it for?
Use it when planning or implementing changes in the Corvus C++ codebase, especially changes involving its agent runtime, CMake integration, offline operation, or MCP support.
Why use it?
It gives coding agents and developers the project context they need before making substantial changes. This reduces misunderstandings about the library's purpose, naming, and intended users.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: mentions CLAUDE.md.

This is Ambar-Gupta22/corvus's own configuration. It tells Claude Code how to work on corvus 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 corvus configures →

not rated 5repo changed 29d ago C scan 5,806 tokens original MIT
View source ↗ Ambar-Gupta22/corvus
Reuse

Borrowing it

Nothing to install: this file belongs to Ambar-Gupta22/corvus. 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/Ambar-Gupta22/corvus/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/Ambar-Gupta22/corvus

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 corvus CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/ambar-gupta22/corvus/claude-md/github.svg)](https://agentmods.dev/instructions/ambar-gupta22/corvus/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/ambar-gupta22/corvus/claude-md"><img src="https://agentmods.dev/badge/instructions/ambar-gupta22/corvus/claude-md/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 corvus CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/ambar-gupta22/corvus/claude-md"><img src="https://agentmods.dev/badge/instructions/ambar-gupta22/corvus/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 5,806 This file is loaded in full into every session.
When invoked 5,806 The same file — it is already loaded in full.
Security scan C 1 finding. 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 $0.05806 $0.05806
Opus 5.5 $0.02322 $0.02322
Sonnet 5.5 $0.01161 $0.01161
Haiku 4.5 $0.00581 $0.00581

Measured 29d ago against content hash b27bc6078855, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade C, and why

corvus CLAUDE.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 29d 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.

Cloud metadata endpointhighServer-side request forgery

One request to 169.254.169.254 can return temporary IAM credentials.

9. **Built-in tools — phasing + safety.** Ship by dependency + blast radius, not all at once: **Calculator + guarded HttpRequest** (Phase 1), **jailed File I/O** (Phase 2), **guarded Shell** (Phase 4, with RBAC). **WebSe
CLAUDE.md · 152 lines

How it starts

The opening of the file, as written. The whole thing — 152 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md — corvus

Guidance for AI assistants (and humans) working in this repo. Read this first.

What this project is

corvus — an in-process, offline-capable, low-latency AI agent runtime for C++. Think "LangChain, but native C++17" — except the real moat is local-first + embeddable + MCP-native, not API mimicry.

Positioning (lead with the capability, not the imitation): An agent runtime that runs where Python can't — ROS2 nodes, game threads, drones, trading loops — and speaks MCP so it works with the existing tool ecosystem on day one.

  • corvus is the working name (namespace corvus, #include <corvus/corvus.h>, CMake target corvus::corvus). Final public name still TBD.
  • "Jarvis" is reserved for a later demo assistant built on top — NOT the library.

Goals

  1. Genuinely useful, elite (top-1%) open-source contribution for C++ developers who have no LangChain equivalent.
  2. Serve: ROS2 robotics, game dev (Unreal), edge/embedded (RPi/Jetson/drones), HFT/low-latency, systems programmers.
  3. Earn adoption via a real problem solved obviously: 12-line quickstart, offline demos, MCP ecosystem, clean CMake integration.

Three differentiators (folded into the design)

  1. MCP-native client — use thousands of existing MCP servers as tools; solves ecosystem cold-start.
  2. Native tool-calling + GBNF — provider tool-use JSON schemas for cloud; grammar-constrained JSON for local llama.cpp models. (ReAct is a fallback, not the default.)
  3. Async + streaming + cancellation — non-blocking runAsync, token streaming, CancelToken. Required for ROS2/game targets (a blocking loop would stall them).

Source-of-truth docs (read before large changes)

Read the full file on GitHub · 152 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. 29d ago Changed · +418 tokens per session b27bc6078855
  2. 1mo ago Changed · +3 lines · +270 tokens per session 279446716f45
  3. 1mo ago Changed · +7 lines · +865 tokens per session b13c72feadfa
  4. 2mo ago First seen · 142 lines · 4,253 tokens per session scan C 6d0a06e89fb8

Subscribe to this mod's changes

corvus CLAUDE.md is an instructions file published in the GitHub repository Ambar-Gupta22/corvus (5 stars, last pushed 29d ago), licensed MIT. It adds 5,806 tokens to every session, about $0.0232 per session on Opus 5.5. A static security scan graded it C with 1 finding (cloud metadata endpoint). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Building agentsC and C++Languagesagentagent-frameworkaiai-agentcpp17
Related

corvus CLAUDE.md compared with instructions from other repositories

corvus CLAUDE.md next to 6 instructions from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
corvus CLAUDE.mdAmbar-Gupta22/corvusInstructions file529d agoMITC5,806Claude Code
vscode buildNext.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA9,719GitHub Copilot
next.js AGENTS.mdvercel/next.jsInstructions file143kyesterdayMITA7,461Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchainInstructions file149kyesterdayMITA4,680Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA5,516GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA3,093GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harnessInstructions file274k2d agoMITA4,313Codex, OpenCode