karpathy-pocock-vectorial-agent-contract
NewAGENTS.md / CLAUDE.md behavioral contract for coding agents, Claude Code, Cursor, Codex, and LLM software engineering. Turn skills into instincts.
Summary
md files, turning explicit instructions into instinctive agent behavior.
- It helps developers enforce consistent coding patterns, tool usage, and decision-making protocols across Claude Code, Cursor, Codex, and other LLM-based software engineering tools.
Install & Usage
mkdir -p .claude/agentsAdd the configuration to .claude/agents/karpathy-pocock-vectorial-agent-contract.md
@karpathy-pocock-vectorial-agent-contractUse Cases
Usage Examples
/karpathy-pocock-vectorial-agent-contract Initialize a behavioral contract for this project with standard coding conventions and tool usage rules.
Create AGENTS.md and CLAUDE.md files that enforce using only functional components in React and prohibit class components.
Update the agent contract to require all database migrations to be reviewed by a senior developer before execution.
Security Audits
Frequently Asked Questions
What is karpathy-pocock-vectorial-agent-contract?
This skill implements a behavioral contract for coding agents using AGENTS.md and CLAUDE.md files, turning explicit instructions into instinctive agent behavior. It helps developers enforce consistent coding patterns, tool usage, and decision-making protocols across Claude Code, Cursor, Codex, and other LLM-based software engineering tools.
How to install karpathy-pocock-vectorial-agent-contract?
To install karpathy-pocock-vectorial-agent-contract: create the agents directory (mkdir -p .claude/agents), then add the config to .claude/agents/karpathy-pocock-vectorial-agent-contract.md. Finally, @karpathy-pocock-vectorial-agent-contract in Claude Code.
What is karpathy-pocock-vectorial-agent-contract best for?
karpathy-pocock-vectorial-agent-contract is a agent categorized under General. It is designed for: agent. Created by ramsani.
What can I use karpathy-pocock-vectorial-agent-contract for?
karpathy-pocock-vectorial-agent-contract is useful for: Establish a shared behavioral contract for all coding agents in a monorepo to ensure consistent code style and tool usage.; Enforce a specific decision-making protocol for refactoring legacy code across multiple LLM tools.; Standardize agent responses to security vulnerabilities by defining a clear escalation and remediation workflow.; Automate the creation of AGENTS.md and CLAUDE.md files for new projects to bootstrap agent alignment.; Prevent agents from using deprecated APIs by codifying allowed libraries and patterns in the contract.; Coordinate multi-agent code reviews by defining review criteria and approval gates in the behavioral contract..