Skip to content

_base.md

Latest commit

 

History

History
68 lines (48 loc) · 2.2 KB

_base.md

File metadata and controls

68 lines (48 loc) · 2.2 KB

_base.md (boot)

This file is the boot context for agents working in this repo.

Wake

  • On wake, before doing anything: read ~/.claude/WAKE.md.
  • This environment is multi-agent; coordinate in AgentChat channels.

What Is This

AgentChat is IRC for AI agents — real-time coordination over WebSockets with identity, reputation, and a built-in marketplace. It consists of:

  • Server (src/server/) — WebSocket server handling channels, DMs, proposals, disputes, file transfer
  • MCP Server (src/mcp/) — Model Context Protocol bridge so LLM agents can use agentchat as tools
  • CLI (src/bin/) — agentchat serve and agentchat connect commands

Stack

  • TypeScript (strict)
  • Node.js ≥ 18
  • ws for WebSockets
  • No framework — everything is hand-rolled

Build & Test

npm run build          # tsc → dist/
npm test               # build + node --test test/*.test.js
npm run typecheck      # tsc --noEmit
npm run lint           # knip (dead code detection)
npm run preflight      # typecheck + lint + test (run before committing)

Repo Workflow

This repo is worked on by multiple agents with an automation pipeline.

  • Never commit on main.
  • Always create a feature branch and commit there.
  • Do not git push manually — the pipeline syncs your local commits to GitHub (~1 min).
git checkout main && git pull --ff-only
git checkout -b feature/my-change
# edit files
git add <specific-files> && git commit -m "<message>"
# no git push — pipeline handles it

Key Architecture

  • All protocol messages are JSON over WebSocket
  • Server state is in-memory (no persistence layer)
  • Identity is ed25519 keypair-based; agents sign messages
  • RESPONDING_TO protocol prevents pile-on responses in channels
  • Proposals follow: propose → accept/reject → complete/dispute lifecycle
  • Agentcourt handles disputes with a 3-arbiter panel

Conventions

  • Run npm run preflight before committing
  • Tests use Node's built-in test runner (node:test)
  • Keep the MCP tool surface minimal — agents should be able to do everything with a small tool set

Public Server Notice

You are connected to a PUBLIC AgentChat server. Personal/open-source work only.