Cross-chain infrastructure built around external bridges inherits every risk those bridges carry. Lithosphere’s MultX handles settlement natively so the stack does not depend on something it cannot control.
Cross-chain bridges have a track record. It is not a good one. The largest single-event losses in blockchain history have not come from compromised execution environments or broken identity systems — they have come from bridges. Lock-and-mint mechanisms with exploitable contracts. Multi-sig setups with insufficient key distribution. Validator sets that turned out to be smaller targets than they looked. The pattern repeats often enough that “bridge risk” has become a standard line item in smart contract audits, a known category of failure that serious infrastructure teams are expected to account for.
The standard response to bridge risk is to pick a bridge that has not been exploited yet, or to spread exposure across multiple bridges in the hope that not all of them fail at once. Neither of these is a solution to the underlying problem. They are risk management strategies for a dependency that the infrastructure still cannot function without. As long as cross-chain settlement runs through an external bridge, the security of that settlement is bounded by the security of the bridge — which is a separate system, maintained by a separate team, with a separate set of assumptions about what constitutes an acceptable attack surface.
Lithosphere’s MultX takes a different position. Cross-chain coordination is built into the Lithosphere stack as a native capability rather than routed through an external bridge after execution completes. When an agent workflow reaches the point where assets or execution need to move across chains, that movement happens within the same architectural environment that handled identity through PPAL and task execution through Lithic — not through a separate system with its own trust assumptions and its own failure modes.
This matters for a reason that goes beyond the obvious risk of bridge exploits. An external bridge does not just introduce a security dependency — it introduces a trust boundary. Whatever guarantees the execution environment provides about how a task ran and what its outcome was, those guarantees do not automatically extend through a bridge to the other side of the cross-chain transaction. The bridge is a handoff point where the context established by the execution environment exits the system. MultX is not a handoff point. It is the same system, operating under the same trust model, applied to the cross-chain step rather than treating that step as someone else’s responsibility.
There is also an operational argument. External bridges go down. They get paused for security reviews. They have congestion periods where transactions queue or fail. Any of these interruptions creates a situation where an agent workflow has completed its execution steps but cannot settle its outcome — not because anything went wrong inside the execution environment, but because the external settlement layer it depends on is temporarily unavailable. MultX’s native integration means the availability of cross-chain settlement is tied to the availability of the Lithosphere stack itself, not to a third party’s uptime and maintenance schedule.
The argument for external bridges has always been pragmatic: building cross-chain settlement natively is hard, and using an established bridge is faster to ship. That argument holds up as long as the risk is acceptable. As agent workflows grow in complexity and value, the risk profile of depending on an external bridge for settlement changes. What is an acceptable trade-off for a simple token transfer is a more serious concern for a multi-step agent workflow where settlement is the final step of a process that may have taken considerable time and cost to reach. Lithosphere built MultX natively precisely because that concern is not hypothetical — it is the direction the industry is moving, and it is better to have addressed it in the architecture than to retrofit a solution after the problem has already materialized.
