judges

Evaluates how accurately an implementation plan accounts for existing code — correctly identifying what to modify vs create, avoiding reimplementation, and finding the right integration points.

Evaluates file and folder structure organization from implementation plans

Evaluates whether an implementation plan is grounded in codebase reality by comparing plan claims against the investigation log. Detects hallucinated file paths, nonexistent modules, and fabricated APIs.

Orchestrates context compression for judge evaluation by determining artifact lists per type, allocating token budgets, and delegating compression

Evaluates whether an implementation plan follows the conventions, patterns, and style found in the actual codebase, as documented in the investigation log.

Evaluates code implementation adherence to custom best practices documents

Evaluates implementation plans for DRY (Don't Repeat Yourself) violations

Evaluates whether an implementation plan addresses the core business/functional goals expressed in the PRD

Evaluates implementation plans for KISS (Keep It Simple) violations

Structural completeness auditor for draft PRDs

Evaluates PRD dependency completeness and integration risk

Evaluates PRD scope discipline and hypothesis traceability

Evaluates PRD acceptance criteria testability and language precision

Evaluates implementation plan readability with focus on clarity, structure, and template adherence

Evaluates code implementation adherence to SOLID Interface Segregation Principle (ISP) and Dependency Inversion Principle (DIP)

Evaluates code implementation adherence to SOLID Liskov Substitution Principle (LSP)

Evaluates code implementation adherence to SOLID Open/Closed Principle (OCP)

Evaluates implementation plans for SSOT (Single Source of Truth) violations

Evaluates technical accuracy of AI assistant responses including API usage, language features, and algorithmic concepts

Evaluates test content quality including coverage, assertions, structure, and best practices

Evaluates whether implementation plan verbosity is appropriately calibrated to problem complexity

Compresses artifacts for judge evaluation. Reads a single raw artifact, applies tiered summarization within a token budget, and returns compacted content with metadata. Isolation via forked context prevents pollution of agent context

Check for a cached plan-evaluation.json result before launching the plan-evaluator agent. This skill should be used in Phase 1.3 (Simple Mode Evaluation) of the orchestrator prompt. Triggers on: entering Phase 1.3, checking simple mode, evaluating plan complexity. Returns EVAL_CACHE_HIT with cached values or EVAL_CACHE_MISS signaling re-evaluation is needed.

Orchestrate parallel judge agent execution, aggregate CaseScore results, write plan-judges.json, code-judges.json, or prd-judges.json, and validate output. Supports evaluating implementation plans (16 judges), code artifacts (11 judges), or PRD artifacts (4 judges) via --artifact-type parameter.