Skip to content

Latest commit

 

History

History
194 lines (143 loc) · 4.83 KB

README.md

File metadata and controls

194 lines (143 loc) · 4.83 KB


 

listen started sa Solana Swiss-Knife toolkit for algorithmic trading, its mission is to become the go-to framework for AI portfolio management agents

Listen Architecture

graph TB
    subgraph "Rig Agent Kit by Listen"
        RAK[RIG Agent Kit]
        RAK_MT[Multi-tenant Stream Manager]
        RAK_WALLET[Delegated Wallet Manager]
        RAK --> RAK_MT
        RAK --> RAK_WALLET
    end

    subgraph "Listen Trading Engine"
        LTE[Trading Engine]
        ORDER_COL[Order Collector]
        PIPE[Pipeline Executor]
        EXEC[Order Executor]

        ORDER_COL --> PIPE
        PIPE --> EXEC
        LTE --> ORDER_COL
    end

    subgraph "Listen Data Service"
        LDS[Data Service]
        SUB[Substreams Indexer]
        DB[(Clickhouse OLAP)]
        PRICE[Price Stream]

        SUB -->|Index Solana Slots| DB
        LDS --> PRICE
        LDS --> DB
    end

    %% External Systems
    MOBILE[Mobile App]
    CHAIN((Blockchain))
    PRIVY((Privy))
    WALLET[(Solana/EVM Wallets)]

    %% Connections
    RAK -->|Tool Calls| CHAIN
    RAK -->|Execute Trades| LTE
    LDS -->|Pricing Updates| LTE
    LDS -->|Enriched Data| MOBILE
    MOBILE -->|User Intents| RAK
    LTE -->|Sign & Send Tx| CHAIN
    DB -->|Query Data| RAK
    RAK_WALLET -->|Integration| PRIVY
    PRIVY --> WALLET
Loading

Features

  • 🔍 Real-time transaction monitoring
  • 💱 Multi-DEX swap execution (Pump.fun, Jupiter V6 API or Raydium)
  • 🚀 Blazingly fast transactions thanks to Jito MEV bundles
  • 📊 Price tracking and metrics
  • 🧰 Token management utilities
  • 📈 Performance monitoring with Prometheus integration

And more!

It works plug'n'play with $arc rig framework framework allowing AI Agents interact with the Solana blockchain, see example: src/agent.rs and the output image.

For complete rundown of features, check out the CLI output of cargo run or the documentation.

Quickstart

To play around with listen-rs, you can use the UI

Fill in the .env.example and ./dashboard/.env.example, copy over to .env and ./dashboard/.env.example, then

docker compose up

You can then access the dashboard over http://localhost:4173

Warning

listen-rs is undergoing rapid iterations, some things might not work and there could be breaking changes

Requirements

  1. System Dependencies

    • Rust (with nightly toolchain)
    • protoc
    • build-essential
    • pkg-config
    • libssl-dev

  2. Configuration

    • Copy .env.example to .env
    • Set up auth.json for JITO authentication (optional, gRPC HTTP/2.0 searcher client)
    • Populate fund.json

Both keypairs are in solana-keygen format, array of 64 bytes, 32 bytes private key and 32 bytes public key.

Quick Start

# Install dependencies
sudo apt install protoc build-essential pkg-config libssl-dev

# Build
cargo build --release

# Run services
./run-systemd-services.sh

Usage Examples

Transaction Monitoring

cargo run -- listen \
  --worker-count [COUNT] \
  --buffer-size [SIZE]

Token Swapping

cargo run -- swap \
  --input-mint sol \
  --output-mint EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v \
  --amount 10000000

Warning

Default configuration is set for mainnet with small transactions. Ensure proper configuration for testnet usage and carefully review code before execution.

Metrics and Monitoring

Listen includes built-in metrics exposed at localhost:3030/metrics. To visualize:

  1. Start Prometheus:
prometheus --config=prometheus.yml
  1. Access metrics at localhost:3030/metrics

Grafana should show something like this

image

Advanced Usage

Swap Profiling

The stackcollapse.pl can be installed through

gh repo clone brendangregg/FlameGraph && \
  sudo cp FlameGraph/stackcollapse.pl /usr/local/bin && \
  sudo cp FlameGraph/flamegraph.pl /usr/local/bin

Profile swap performance using DTrace to produce a flamegraph:

./hack/profile-swap.sh
image