The Risk Management agent has no check for whether the target exchange is actually open before the Portfolio Manager approves execution. The framework can submit orders during market closures, DST transition windows, or exchange halts — there's no circuit breaker for this.
I put together an integration guide showing how to add a signed pre-trade check as a single function call in the risk management step:
https://headlessoracle.com/docs/integrations/trading-agents
It returns a cryptographically signed receipt (OPEN/CLOSED/HALTED/UNKNOWN). If UNKNOWN, the safe default is halt — don't let the trade proceed. Takes about 5 lines to wire in.
Happy to send a PR if this is useful to the maintainers.
The Risk Management agent has no check for whether the target exchange is actually open before the Portfolio Manager approves execution. The framework can submit orders during market closures, DST transition windows, or exchange halts — there's no circuit breaker for this.
I put together an integration guide showing how to add a signed pre-trade check as a single function call in the risk management step:
https://headlessoracle.com/docs/integrations/trading-agents
It returns a cryptographically signed receipt (OPEN/CLOSED/HALTED/UNKNOWN). If UNKNOWN, the safe default is halt — don't let the trade proceed. Takes about 5 lines to wire in.
Happy to send a PR if this is useful to the maintainers.