npx claudepluginhub athola/claude-night-market --plugin leylineThis skill uses the workspace's default tool permissions.
A decision framework for agent orchestration based on Liu et al.,
Enforces C++ Core Guidelines for writing, reviewing, and refactoring modern C++ code (C++17+), promoting RAII, immutability, type safety, and idiomatic practices.
Provides patterns for shared UI in Compose Multiplatform across Android, iOS, Desktop, and Web: state management with ViewModels/StateFlow, navigation, theming, and performance.
Implements Playwright E2E testing patterns: Page Object Model, test organization, configuration, reporters, artifacts, and CI/CD integration for stable suites.
A decision framework for agent orchestration based on Liu et al., "Utility-Guided Agent Orchestration for Efficient LLM Tool Use" (arXiv:2603.19896). Each candidate action is scored by subtracting weighted costs from expected gain, producing a single utility value that guides action selection. The framework prevents over-calling tools and premature stopping by making both errors costly. Utility range is [-2.3, 1.0].
A = {respond, retrieve, tool_call, verify, delegate, stop}
| Action | Description |
|---|---|
| respond | Emit a final answer from current context |
| retrieve | Fetch additional information (search, read, lookup) |
| tool_call | Execute a tool (code runner, API, file write) |
| verify | Check a prior result for correctness or completeness |
| delegate | Spawn a sub-agent or hand off to a specialist |
| stop | Terminate the loop and return current state |
U(a | s_t) = Gain(a | s_t)
- λ₁ · StepCost(a | s_t)
- λ₂ · Uncertainty(a | s_t)
- λ₃ · Redundancy(a | s_t)
| Parameter | Default | Rationale |
|---|---|---|
| λ₁ | 1.0 | Cost baseline; all other weights relative to this |
| λ₂ | 0.5 | Weak empirical correlation with outcome (r=0.0131) |
| λ₃ | 0.8 | Redundancy pruning yields ~10% token savings |
Utility range: [-2.3, 1.0]. Positive values indicate the action is worth taking. Values below the floor (-0.5 default) indicate the action should be skipped.
Stop the loop when any of the following is true:
stopstop actions score below the floor (default: -0.5)High-gain override: If Gain >= 0.7 for any action, condition
(c) may be overridden.
Document the override and the gain value in your reasoning trace.
Minimal 4-step advisory pattern:
modules/state-builder.mdA per
modules/action-selector.mdU(a | s_t), subject to termination conditionsmodules/state-builder.md -- how to
populate s_t from task contextmodules/gain.md -- estimating expected information
or progress gainmodules/step-cost.md -- token, latency, and
monetary cost tablesmodules/uncertainty.md -- confidence
estimation and calibrationmodules/redundancy.md -- detecting duplicate
or low-delta actionsmodules/action-selector.md -- scoring
loop and tie-breaking rulesmodules/integration.md -- wiring utility
scoring into existing orchestration loops