stacks-alpha-engine
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from blockchain APIs and user-supplied inputs to generate DeFi transaction instructions, creating an indirect prompt injection surface.
- Ingestion points: The
stacks-alpha-engine.tsscript fetches real-time token prices, liquidity pool metrics, and wallet positions from the Hiro, Tenero, and Bitflow APIs. User inputs include wallet addresses and transaction amounts. - Boundary markers: The
AGENT.mdinstructions establish a strict decision pipeline (Scout -> Reserve -> Guardian -> Executor). All write commands default to a dry-run preview mode and require an explicit--confirmflag before generating executable instructions. - Capability inventory: The tool generates transaction instructions for supply, borrow, and liquidity operations across the Zest, Hermetica, Granite, and Bitflow (HODLMM) protocols. It does not execute these transactions directly but outputs them for the agent runtime to handle.
- Sanitization: The skill validates Stacks wallet addresses via Regular Expression (
/^SP[A-Z0-9]{30,}$/i), parses amounts as integers, and applies market safety thresholds including 24h volume floors and slippage caps. - [EXTERNAL_DOWNLOADS]: The skill fetches data and configuration from external blockchain infrastructure providers and depends on standard Node.js packages.
- Retrieves wallet balances and on-chain state from the Hiro Stacks API (
api.mainnet.hiro.so). - Fetches market pricing from the Tenero Price Oracle (
api.tenero.io). - Accesses decentralized exchange data from the Bitflow API (
bff.bitflowapis.finance). - Verifies Bitcoin reserve integrity using the Mempool API (
mempool.space/api). - Dependencies include standard libraries:
commanderfor CLI management andtiny-secp256k1for cryptographic operations.
Audit Metadata