pp-defillama

Pass

Audited by Gen Agent Trust Hub on Aug 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructions direct the agent to install the defillama-pp-cli tool using npx or go install from the vendor's repositories.
  • [REMOTE_CODE_EXECUTION]: The installation of the CLI and MCP server involves downloading and executing code from external package registries (npm) and GitHub.
  • [COMMAND_EXECUTION]: The skill operates by executing the defillama-pp-cli binary via shell commands.
  • [DATA_EXFILTRATION]: The CLI features a --deliver webhook: flag, which enables routing command output to an external URL. This is a powerful data routing feature that could be misused for exfiltration.
  • [PROMPT_INJECTION]: The skill processes data from the DefiLlama API, creating a surface for indirect prompt injection.
  • Ingestion points: API data fetched by defillama-pp-cli (SKILL.md).
  • Boundary markers: None present.
  • Capability inventory: Includes shell command execution and file reading (SKILL.md).
  • Sanitization: The skill uses structured JSON output in agent mode to mitigate unstructured data risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 18, 2026, 05:33 AM
Security Audit — agent-trust-hub — pp-defillama