opinion
Pass
Audited by Gen Agent Trust Hub on Feb 20, 2026
Risk Level: SAFECREDENTIALS_UNSAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [CREDENTIALS_UNSAFE] (LOW): The skill instructions (SKILL.md) direct the user to export sensitive credentials, including an API key and a wallet private key (
OPINION_PRIVATE_KEY), to environment variables. While these are placeholders in the documentation, this pattern encourages a practice that can lead to credential exposure in shell history or process listings. - [INDIRECT_PROMPT_INJECTION] (LOW): The skill identifies a surface for indirect prompt injection.
- Ingestion points: Market data, titles, and descriptions are fetched from the external Opinion.trade API via the
/op marketsand/op marketcommands. - Boundary markers: There are no instructions or delimiters defined to prevent the agent from obeying instructions that might be embedded in market descriptions.
- Capability inventory: The skill possesses the capability to perform financial transactions (buy/sell orders) on the BNB Chain.
- Sanitization: No sanitization or validation of the external market data is described.
- [NO_CODE] (SAFE): The provided skill package does not contain any executable scripts, binaries, or logic files, which significantly reduces the risk of direct malicious execution from the skill itself.
Audit Metadata