Security audit
Security checks for vulnerabilities and agentic risk
The skill openly documents paid API services with spending controls, though users should review its full service list and consider what data they share.
Review the full endpoint catalog and agree on a total budget before paying. Share only content you are comfortable sending to Blackwood State and its processing providers. Use a lightly funded dedicated wallet and enable optional private-key signing only if needed.
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.
The skill instructs the agent to POST request bodies to externally hosted endpoints, including for text processing and chat tasks, but never warns that supplied user content leaves the local environment. The payment approval requirement discloses spending, not disclosure of potentially sensitive input to Blackwood State and its processing providers.
The manifest at L003 limits the skill to onchain data, ECB exchange rates, and quick AI text utilities. However, L103–L123 supplies callable POST endpoints for weather forecasts, airport delays, time-zone conversion, image and speech generation, QR codes, and random numbers. These are distinct paid services rather than implementation details of the declared functions, expanding what the agent can invoke and spend funds on.
No suspicious patterns detected.
