Shimmy$SHIMMYv1.0.0
pump.funDexScreenerTwitterCommunityAboutDocs
Model Fleet6 models
SETUPMac Mini · M-series · Local inference
4 steps → autonomous
1Install runtime
copy
$ curl -fsSL https://shimmy.sh/install | sh
Single binary, no Python, no Docker
2Pull a model
copy
$ shimmy pull llama-3.1-70b --quant q4
GGUF / SafeTensors, auto-quantize to fit VRAM
3Start serving
copy
$ shimmy serve --port 8080 --api openai
OpenAI-compatible API on your local network
4Deploy agent
copy
$ shimmy agent start --strategy sniper
Autonomous trading via on-chain skills
AGENT SKILLS6 available
▸
token-scanner
Real-time Solana token discovery via DexScreener
▸
liquidity-probe
Deep liquidity & holder distribution analysis
▸
sniper-entry
Sub-second buy execution on new pairs
▸
exit-engine
Auto-extract on first pump, configurable targets
▸
rug-detector
Contract analysis, mint authority, LP lock checks
▸
wallet-cluster
Cross-reference wallet patterns for insider detection
shimmy v1.0.0 · Apple Silicon optimizedzero cloud fees · your hardware · your edge
Modelllama-3.1-70b
Latency0ms
Throughput0 tok/s
Total Tokens0
IDLE
Initializing token analysis engine...
shimmy inference · on-chain token analysismodel: llama-3.1-70b · rust inference engine
Initializing token scanner...
Fleet operational · Click any model to hot-swap
shimmy v1.0.0 · rust · apache-2.0