Hook Claude Code
Runs before the agent uses a tool for Read, read, read_file, file_read, Grep, grep, Search, search and code_search tool calls, running warden hook redirect (2 commands). From rynald0cst0ltziam/Warden-AI.
Settings file Claude Code
Agent settings declaring 1 hook event (PreToolUse).
Cursor rule Cursor
This project uses Warden, an MCP server that manages context across eight layers: code intelligence, context selection, tool output pruning, agent memory, response compression, file compression, MCP description compression, and session continuity. It proves every optimization is safe via a shadow-mode eval gate.
Instructions file GitHub Copilot
Copilot instructions for rynald0cst0ltziam/Warden-AI, covering warden — context governance and verification layer, session start (important — do this first), layer 1: before starting work — context selection, layer 2: during work — tool output pruning and enforcement hooks (automatic).
Instructions file CodexOpenCode
AGENTS.md instructions for rynald0cst0ltziam/Warden-AI, covering agents.md — warden, what this is, build / test / verify commands, architecture (where things live) and non-negotiable rules.
Instructions file
Claude Code instructions for rynald0cst0ltziam/Warden-AI, covering warden — context governance and verification layer, session start (important — do this first), layer 1: before starting work — context selection, layer 2: during work — tool output pruning and enforcement hooks (automatic).
Instructions file Gemini CLI
Gemini CLI instructions for rynald0cst0ltziam/Warden-AI, covering warden — context governance and verification layer, session start (important — do this first), layer 1: before starting work — context selection, layer 2: during work — tool output pruning and enforcement hooks (automatic).