skills/openclaw/skills/dexter/Gen Agent Trust Hub

dexter

Warn

Audited by Gen Agent Trust Hub on Feb 24, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Clones source code from a third-party GitHub repository (github.com/virattt/dexter.git) that is not listed as a trusted vendor.\n- [REMOTE_CODE_EXECUTION]: Executes 'bun install' and runs the application using 'bun', which results in the execution of unvetted third-party code.\n- [COMMAND_EXECUTION]: Uses 'sed' for automated source code modification and dynamically generates TypeScript wrapper scripts (query.ts) for execution.\n- [PROMPT_INJECTION]: The agent processes untrusted data from web searches and financial filings, creating a surface for indirect prompt injection.\n
  • Ingestion points: Web search results via Tavily and SEC filing data via Financial Datasets API.\n
  • Boundary markers: No explicit markers or 'ignore' instructions are provided in the query wrapper.\n
  • Capability inventory: File system writing, network API calls, and shell execution via bun.\n
  • Sanitization: No evidence of sanitization or validation of the external content before synthesis.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 24, 2026, 07:56 AM