caribou-strategy
Warn
Audited by Snyk on Jun 28, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.90). The README explicitly instructs fetching and then executing remote code needed at runtime (npx install from https://github.com/Senpi-ai/senpi-skills and curl of raw files from https://raw.githubusercontent.com/Senpi-ai/senpi-skills/main/caribou/$f), so these URLs supply required code that will be executed by the producer daemon and thus directly affect runtime behavior.
MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).
- Direct money access detected (high risk: 1.00). The skill is explicitly a trading/managed-futures system designed to open and manage financial positions. It references wallets, opening LONGS and SHORTS, margin/leverage formulas, order types (FEE_OPTIMIZED_LIMIT with taker fallback), and explicitly named execution actions (create_position, close_position, edit_position, ratchet_stop_, cancel_order, strategy_close). Those are direct market-execution capabilities (placing/editing/canceling orders and managing wallet-position state), not generic tooling. Even though user sessions are declared "READ-ONLY", the presence of those execution APIs and wallet bindings shows the skill is specifically built to move/manage money/assets.
Issues (2)
W012
MEDIUMUnverifiable external dependency detected (runtime URL that controls agent).
W009
MEDIUMDirect money access capability detected (payment gateways, crypto, banking).
Audit Metadata