beaver-strategy

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to download scripts and configuration files from the official Senpi GitHub repository (github.com/Senpi-ai/senpi-skills) during installation.\n- [COMMAND_EXECUTION]: The README guide details steps to register plugins, install dependencies using npx, and execute the producer script using python3. These are routine setup operations for this strategy type.\n- [PROMPT_INJECTION]: The skill ingests untrusted market data and sentiment metrics which are interpolated into LLM prompts for decision making.\n
  • Ingestion points: The beaver_entry action in runtime.yaml processes signal data from the beaver_signals scanner.\n
  • Boundary markers: The prompt uses a SIGNAL: header to separate data from instructions.\n
  • Capability inventory: The skill can execute trading actions such as OPEN_POSITION.\n
  • Sanitization: No explicit sanitization or filtering of external data content is implemented prior to its use in the decision prompt.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 08:48 PM
Security Audit — agent-trust-hub — beaver-strategy