crypto-bd-agent
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill architecture specifies ingesting untrusted data from sources like Twitter/X, protocol forums, and general web scraping (via Firecrawl) for automated token evaluation and scoring. There are no specified prompt boundary markers or sanitization procedures for this external content before it is processed by the LLM cascade, creating a surface for indirect prompt injection. The skill also generates outreach drafts based on this data, which could be influenced by malicious content.
- [NO_CODE]: The skill consists exclusively of markdown documentation and YAML metadata. It does not include any scripts, binaries, or automated shell commands that would execute code in the agent's environment.
Audit Metadata