agents▌

Create and maintain minimal, high-signal agent documentation under 60 lines. \n \n Enforces research-backed best practices for agent-facing docs; instruction quality degrades with length \n Requires three core sections: Package Manager, File-Scoped Commands (per-file test/lint/typecheck), and Commit Attribution with agent identity \n Analyzes project structure (lock files, linter configs, CI commands, monorepo indicators) to determine what belongs in the file \n Uses headers and bullets only; re

Persistent, stateful AI agents with thread management, tool integration, streaming, and RAG on Convex. \n \n Thread management for multi-turn conversations with automatic persistence across restarts and real-time streaming responses to clients \n Tool integration allowing agents to execute Convex functions as callable tools for knowledge search, task creation, and external API calls \n Built-in vector search and RAG patterns for embedding documents and retrieving relevant context to augment agen

Foundation framework for building multi-step agents with built-in planning, memory, and skill delegation. \n \n Provides six core middleware options: task planning, filesystem context management, subagent delegation, persistent memory, human approval workflows, and on-demand skill loading \n Includes three always-present built-in tools: write_todos for task tracking, filesystem operations ( ls , read_file , write_file , edit_file , glob , grep ), and task for spawning specialized subagents \n Su

Guide for creating AGENTS.md files for project-specific inline rules.

When launching parallel background agents, do NOT poll with sleep loops.

Orchestration through Claude Code's built-in Agent Tool

When launching multiple agents in parallel, follow this pattern to avoid context bloat.

When you have multiple unrelated failures (different test files, different subsystems, different bugs), investigating them sequentially wastes time. Each investigation is independent and can happen in parallel.

The fundamental architecture of computer use agents: observe screen, reason about next action, execute action, repeat. This loop integrates vision models with action execution through an iterative pipeline.

Build stateful AI agents on Cloudflare Workers with WebSockets, persistent state, scheduling, and multi-agent coordination. \n \n WebSocket-based real-time communication with automatic state synchronization across clients and devices; resumable streaming persists across disconnects and page refreshes \n Built-in SQLite storage (up to 1GB per agent), task scheduling with cron expressions, and Durable Objects for globally unique, persistent agent instances \n Multi-agent coordination via routeAgen