okx-activity
Pass
Audited by Gen Agent Trust Hub on Aug 7, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXFILTRATION]: The skill implements mandatory masking of sensitive identifiers (OKX UID) in console output and explicitly forbids logging or passing JWT tokens in CLI flags. These measures effectively prevent accidental data exposure.
- [COMMAND_EXECUTION]: The skill executes
onchainosCLI commands to list agents and perform registration. All commands are purpose-built for the hackathon registration flow and include standard error handling viajqto ensure platform stability. - [EXTERNAL_DOWNLOADS]: The skill includes a 'pre-flight' check that may initiate a CLI upgrade from the official OKX GitHub repository if version drift is detected. This is a standard maintenance procedure for vendor-provided tools.
- [PROMPT_INJECTION]: The skill instructions include 'Wrong-skill guards' and strict confirmation requirements ('Gate B') to ensure the agent does not perform irreversible registration actions without explicit, numbered user confirmation. These guards prevent the agent from misinterpreting ambiguous user requests.
Audit Metadata