Skip to content

README.md

Latest commit

 

History

History
67 lines (51 loc) · 2.73 KB

README.md

File metadata and controls

67 lines (51 loc) · 2.73 KB
Raw

Trenchsi Core

Trenchsi shrimp avatar

This directory contains the Trenchsi agent core: the Go CLI, startup launcher, web console backend, workspace scaffold, provider stack, tool system, automation features, and multi-channel integrations, with a strong focus on memecoin trenching and trading workflows.

For the main project overview and quick start, go to the repository README.

Start Here

go build -o trenchsi ./cmd/trenchlaw
./trenchsi onboard
./trenchsi

If you want the terminal agent directly, use:

./trenchsi agent

For Solana trading, use the dedicated runner:

./trenchsi run --strategy solana-trench --dry-run

That path reads its config from config/config.json plus the env vars in .env.example, loads the Solana wallet from SOLANA_PRIVATE_KEY, scans Jupiter token candidates, and applies the conservative trenching filters before any live swap is attempted.

Typical first-run flow:

./trenchsi auth login
./trenchsi model
./trenchsi status

Core Features

  • Built for memecoin trenching, trade support, and fast operator workflows
  • Interactive startup launcher with terminal, web, and TUI entry points
  • Optional Solana trenching runner with dry-run, paper trading, and live swap support
  • Configurable providers and models with auth flows and provider fallback support
  • Workspace-driven identity, memory, tools, and user customization
  • Skill discovery, installation, and per-workspace skill management
  • Web search, file operations, shell execution, and extensible tool registry
  • MCP integration with deferred tool discovery
  • Session history and persistent memory storage
  • Gateway-based chat integrations across Telegram, Discord, Slack, Matrix, WhatsApp, QQ, WeCom, Feishu, OneBot, and more
  • Agent bindings for routing different channels or contexts to different agents
  • Built-in scheduling and recurring jobs with cron commands
  • Sandbox controls and sensitive data filtering for safer tool execution

Documentation