Mechanism
The effort analysis module is a component of the confidence computation subsystem. It evaluates the projected effort cost of executing the current task along the currently selected execution path. Effort cost is a composite measure that quantifies the total expenditure of computational resources, time, energy, coordination overhead, and cognitive load required to complete the task along a given path. The module exists because capability and effort are different questions. An agent may possess every capability a task requires and still face a path that consumes resources out of proportion to the value of the outcome, or out of proportion to an alternative path that reaches the same or a comparable outcome at lower cost.
Effort cost is therefore structurally distinct from capability sufficiency. Capability sufficiency asks whether the agent can do the task at all. Effort cost asks what the chosen way of doing it will consume. The effort analysis module supplies the second answer and feeds it into the confidence computation as its own input dimension.
Effort Cost Across Candidate Paths
The effort analysis module does not evaluate the selected path in isolation. It computes effort cost for each candidate execution path identified by the forecasting engine's planning graphs. For a given task, the forecasting engine may have identified a plurality of eligible branches, each representing a distinct execution strategy with a distinct projected outcome, projected cost, and projected risk profile. The effort analysis module augments this set by computing, for each eligible branch, a normalized effort metric: the ratio of projected resource expenditure to projected outcome value.
This ratio is what makes the comparison meaningful across paths that differ in both cost and result. A branch with a high effort metric represents high expenditure relative to value, which is an inefficient path. A branch with a low effort metric represents low expenditure relative to value, which is an efficient path. The metric is normalized so that a cheap path producing a weak outcome and an expensive path producing a strong outcome can be ranked against one another rather than compared on raw cost alone.
Effort as a Confidence Input
The effort metric contributes directly to the confidence computation. A high-effort path reduces the agent's confidence even when capability sufficiency, resource availability, and every other confidence input is favorable. The agent's confidence that it can execute well decreases as the projected effort increases, even when the agent's capabilities are theoretically sufficient for the task.
This is the structural point of the module. Without it, an agent that holds the right capabilities would report high confidence regardless of how wastefully it intended to proceed. Coupling the effort metric to confidence means that the cost of the chosen approach, not just the agent's possession of the relevant skills, governs whether execution is warranted.
Path-of-Least-Resistance Computation
The effort analysis module implements a path-of-least-resistance computation: a mechanism by which the agent identifies the execution path that achieves the required task outcome with the minimum effort cost. The computation operates over the set of eligible planning graph branches and ranks them by their normalized effort metric, producing an ordered list from the least-effort path to the most-effort path.
The agent's confidence computation is then evaluated against the least-effort path rather than against the agent's currently selected path. This produces two confidence values for the same task: an as-planned confidence, reflecting the agent's confidence in executing the path it currently intends to take, and an optimized confidence, reflecting the agent's confidence in executing the least-effort path. Holding both values at once is what lets the agent recognize that a better way to reach the same outcome is available.
Path Recommendation
When the optimized confidence exceeds the as-planned confidence by more than a configurable improvement threshold, the effort analysis module generates a path recommendation: a structured suggestion that the agent should consider switching to the lower-effort path. The recommendation is advisory in a precise sense. It does not override the agent's current execution plan. It is presented to the agent's deliberation pipeline as an input, subject to the same intent evaluation, policy checking, and affective modulation that govern all deliberation inputs.
The agent may accept, reject, or defer the path recommendation. The grounds for that decision extend beyond effort cost, because the effort metric does not capture everything that matters. Intent alignment, risk profile, and strategic considerations may all justify keeping a higher-effort path. The module surfaces the efficiency opportunity; it does not force the agent to take it.
Iterative Effort Re-evaluation
Effort analysis does not stop at the pre-execution projection. The module supports iterative effort re-evaluation during execution. As the agent progresses along its selected path, actual resource consumption is compared against projected resource consumption, and the effort metric is updated with the observed data.
The update feeds back into confidence. If actual effort exceeds projected effort, indicating that the path is more costly than anticipated, the effort contribution to confidence increases and confidence decays. If actual effort falls below projected effort, indicating that the path is cheaper than anticipated, the effort contribution decreases, which supports confidence maintenance or recovery. This keeps the effort analysis grounded in observed conditions rather than relying exclusively on pre-execution estimates that the actual execution may contradict.
Task-Class-Aware Weighting
The path-of-least-resistance computation interacts with the task class differentiation applied elsewhere in the confidence governor. It does not optimize naively for resource minimization, because the least-effort path is not always the appropriate path. The module weights its evaluation according to the class of the task in hand.
For terminal tasks, characterized by high irreversibility, the module weights reliability and safety more heavily than effort minimization, preferring a higher-effort path with lower risk over a lower-effort path with higher risk. For exploratory tasks, it weights breadth of exploration more heavily, preferring paths that cover more of the search space even at higher per-unit effort. For generative tasks, it weights creative optionality more heavily, preferring paths that preserve more creative degrees of freedom even at higher nominal effort. This task-class-aware weighting ensures that effort analysis serves execution quality rather than undercutting it.
Disclosure Scope
The effort analysis module, comprising the composite effort cost measure, the normalized effort metric computed per eligible planning graph branch as the ratio of projected resource expenditure to projected outcome value, the contribution of that metric to the confidence computation, the path-of-least-resistance computation that ranks eligible branches and yields paired as-planned and optimized confidence values, the path recommendation generated when the optimized confidence exceeds the as-planned confidence beyond a configurable improvement threshold, the iterative effort re-evaluation that updates the metric from observed consumption during execution, and the task-class-aware weighting across terminal, exploratory, and generative tasks, is disclosed in the cognition filing (U.S. Application No. 19/647,395 and its international counterpart) at Section 5.11. This article describes that disclosed mechanism. The disclosure separates the structural treatment of effort cost as a first-class confidence input from the numerical calibration of the improvement threshold, the metric's normalization, and the per-class weights, which are implementation choices that vary by deployment.
