Mechanism
Capability envelope negotiation is the protocol the system may initiate when a substrate's capability envelope does not fully match an objective's requirements, but one or more dimensions are conditionally satisfiable rather than unsatisfied. A conditionally satisfiable dimension is one in which the substrate's capability currently falls short of the requirement but could be brought into satisfaction, for example through a reconfiguration that provides the required capability. Negotiation is the mechanism that explores the reconfiguration path: instead of rejecting the substrate, the agent and the substrate exchange structured information about what modification would close the gap and at what cost.
The trigger is therefore not a contested claim between peers. It is a per-dimension match result. Capability matching produces, for each dimension, one of three outcomes: satisfied, unsatisfied, or conditionally satisfiable. When at least one dimension is conditionally satisfiable through reconfiguration, the system may open negotiation with that substrate to determine whether the conditional gap is worth closing.
What the Substrate Advertises
During negotiation, the substrate advertises the specific modifications it could make to its own capability envelope. The disclosure gives concrete examples: loading a required inference model, provisioning a GPU partition, allocating reserved memory, or activating a dormant sensor. Each advertised modification is accompanied by the estimated time cost and the estimated resource cost of performing it. The substrate is, in effect, describing the price of becoming capable along the conditionally satisfiable dimension, expressed in time and in resources rather than in an abstract score.
This places the negotiation firmly between an agent and a substrate, not between two competing agents. The substrate is the party that holds and can mutate its own envelope, so it is the party that advertises the available envelope modifications and their costs.
How the Agent Evaluates the Offer
The agent evaluates whether the modification cost is justified relative to the value of executing on this substrate versus rerouting to an alternative substrate whose capability envelope already satisfies the requirements without modification. Rerouting is one of the determinate capability outcomes, so it is always the comparison baseline: if another substrate is already capable, paying the time and resource cost to modify this one may not be warranted.
Three considerations enter the evaluation. The first is the temporal executability window that would result from the modification, computed to include the time to perform the modification itself, so that the agent reasons about when execution would actually become possible rather than only whether it could. The second is the opportunity cost of occupying the substrate's resources during the modification. The third is the agent's affective state: an agent with elevated urgency may reject modification-dependent paths in favor of immediately available alternatives. The affective state thus shapes the negotiation outcome without changing the protocol.
The Capability Acquisition Plan
If the agent elects to proceed with modification, it issues a capability acquisition plan. The acquisition plan is a structured request that authorizes the substrate to modify its envelope, and it is subject to governance approval. The modification is therefore not a side effect the agent triggers unilaterally; it is a governed action that passes through the same approval discipline as other governed mutations before the substrate is permitted to act on its own envelope.
Once authorized, the modification is executed, the capability envelope is updated to reflect the new configuration, and the capability determination is re-evaluated against the updated envelope. The conditionally satisfiable dimension that opened the negotiation is the dimension whose status the re-evaluation is meant to resolve. The complete negotiation transaction is recorded in the agent's lineage.
Re-Evaluation and Closure
Negotiation does not assume success. After the envelope is updated, the capability determination is computed again, which means the post-modification substrate is subject to the same per-dimension matching and the same aggregate composition rule as before. A modification that was expected to bring a dimension into satisfaction is confirmed by re-evaluation rather than presumed. If the re-evaluated determination resolves to structurally possible, execution synthesis can proceed on the now-capable substrate. If it does not, the determinate outcomes remain available, including rerouting to an alternative substrate or treating the dimension as unsatisfied.
The negotiation transaction sits in lineage as a complete record: the conditional gap that prompted it, the advertised modifications and their costs, the agent's evaluation, the issued acquisition plan and its governance approval, the modification, and the re-evaluated determination. A governance auditor can reconstruct why a substrate was modified rather than bypassed, and whether the modification achieved the capability it was authorized to achieve.
Distinction from Resource Negotiation and Contention
The disclosure is explicit that capability envelope negotiation is structurally distinct from two neighboring mechanisms. Governed substrate resource negotiation, in which agents negotiate with substrates and with other agents for processor allocation, memory budget, network bandwidth, and sensor access, occurs before execution and produces binding resource commitments through a declaration, counteroffer, and commitment sequence. Capability envelope negotiation is different in kind: it resolves conditional capability gaps through substrate modification rather than allocating already-available resources.
Contention resolution is different again. Contention is the condition in which multiple agents simultaneously require the same substrate capability or the same temporal executability window, and it is resolved after competing claims arise by evaluating each agent's forecasted executability and rerouting agents that have viable alternatives. Capability envelope negotiation does not arbitrate between competing agents; it asks whether a single substrate can be reconfigured to satisfy a single agent's conditionally satisfiable requirement.
Distinction from Conventional Dispatch
Conventional distributed systems make a capability decision at dispatch time, typically through resource-availability checks or static capability registries, and then treat the inability to execute as a failure requiring error handling, retry logic, or escalation. Such systems have no structured path between can and cannot: a substrate that is not currently capable is simply not selected, even if a modest reconfiguration would make it capable.
Capability envelope negotiation occupies exactly that missing middle. Because the matcher distinguishes conditionally satisfiable from unsatisfied, the system can treat near-misses as negotiable rather than as failures, weigh the time and resource cost of closing the gap against rerouting, obtain governance approval for the modification through a capability acquisition plan, and confirm the result by re-evaluation. The path from conditional gap to capable substrate is a governed, recorded transaction rather than a retry.
Disclosure Scope
Capability envelope negotiation, comprising the initiation of the protocol when a substrate's capability envelope is conditionally satisfiable along one or more dimensions, the substrate's advertisement of specific envelope modifications such as loading an inference model, provisioning a GPU partition, allocating reserved memory, or activating a dormant sensor together with their estimated time and resource costs, the agent's evaluation of modification cost against rerouting in light of the resulting temporal executability window, the opportunity cost of occupying the substrate, and the agent's affective state, the issuance of a governance-approved capability acquisition plan authorizing the substrate to modify its envelope, and the execution of the modification followed by re-evaluation of the capability determination with the complete transaction recorded in the agent's lineage, is disclosed in the cognition filing (U.S. Application No. 19/647,395 and its international counterpart). This article describes that disclosed mechanism and its stated distinction from governed substrate resource negotiation and from contention resolution. Deployment-specific cost models, governance approval policies, and the set of modifications a given substrate can advertise are configuration of the disclosed framework and are not themselves limitations of it.
