At the base layer of any blockchain lies a finite set of transaction types that define what the network can express natively. For Kite Protocol, which is purpose-built for autonomous AI agents rather than human-driven transfers, this layer must remain both extensible and carefully protected. The ability to add new Agent Transaction Types is therefore treated as a governance-sensitive operation, managed directly by KITE stakers through a structured, on-chain decision process.
Agent Transaction Types are foundational protocol primitives. They are not application-level abstractions or optional extensions, but first-class transactions recognized by validators and enforced at the consensus level. These transactions encode agent-specific actions—such as committing to continuous micropayments, binding payments to verifiable computation, or expressing cross-chain intents—in a way that is natively understood by the network. Because each new type expands the protocol’s expressive power, it also expands its attack surface. Governance exists to ensure that this expansion is deliberate, justified, and technically sound.
Kite formalizes this process through Transaction Type Proposals (TTPs). A TTP is a comprehensive, on-chain submission that combines technical specification with economic and security rationale. Proposers must define the transaction’s structure, validation rules, and state interactions, including how it interfaces with agent identities, staking constraints, or reputation systems. Just as importantly, the proposal must articulate a concrete agent use case that cannot be adequately served by existing transaction types. This requirement helps distinguish genuine protocol-level needs from features better handled at the application layer.
Once submitted, a TTP moves through a staged governance pipeline. The initial phase is open technical review, where the broader community—including protocol engineers, researchers, and experienced builders—examines feasibility, composability, and potential failure modes. This stage often results in revisions, clarifications, or even withdrawal if the proposal proves misaligned or overly complex. The emphasis here is on early risk reduction rather than voting momentum.
For proposals that demonstrate clear utility and architectural fit, governance can authorize the use of treasury resources to fund independent security assessments. This step acknowledges that base-layer changes warrant scrutiny beyond informal review. By separating audit funding from final approval, Kite reduces the likelihood that critical flaws surface only after a binding vote has already passed.
The binding decision is made through a stake-based governance vote, typically weighted by lock duration. Tokens committed for longer periods carry greater influence, reflecting a higher level of long-term exposure to protocol outcomes. Given the systemic impact of introducing new transaction types, approval thresholds are intentionally conservative, often requiring a supermajority. This ensures that upgrades reflect broad consensus rather than narrow enthusiasm.
Approval does not equate to immediate activation. Newly accepted transaction types are deployed first to a testnet environment designed to mirror mainnet conditions. This phase allows integrators to experiment, validators to observe performance characteristics, and auditors to conduct additional analysis under real usage patterns. Only after this period, and a final confirmation by governance, is the transaction type enabled on the main network.
This layered process serves multiple objectives simultaneously. It protects protocol integrity by enforcing technical discipline, aligns incentives by placing decision power with economically committed stakeholders, and preserves adaptability by offering a clear path for innovation. Rather than freezing the base layer or centralizing its evolution, Kite embeds change control into governance itself.
In practice, this means Kite’s base layer evolves in response to real agent behavior and developer needs, not speculative roadmaps. As new forms of coordination, computation, and value exchange emerge among autonomous agents, KITE stakers act as curators of the protocol’s expressive capacity. The network remains neither static nor permissive, but carefully extensible—guided by those with both technical insight and long-term stake in its trajectory.