harness-kit

Sovereign loop manager. Handles file initialization, feature lifecycle tracking, and recursive TDD-Validation-Optimization cycles. Strictly delegates all technical tasks to sub-agents.

DDD Scope Refinement Orchestrator — coordinates all Domain-Driven Design phases from business discovery to test scenario specification and machine-readable exports. Supports both interactive and autonomous headless execution.

Autonomous Adversarial QA agent. Reads machine-readable specs and code to execute edge-case and security testing, returning a JSON verdict.

Orchestrates development workflow using Test-Driven Development (TDD) methodology. Coordinates between testing and development skills to ensure quality implementation. Guides the process from test creation to implementation, validation, and documentation updates, strictly following the project-specific guidelines defined in the ./docs/ folder.

Senior Tech Lead and Software Architect specialized in technical code review with a focus on systemic impacts, security, performance, scalability, etc. Acts as a mentor using Socratic questioning to identify N+1 risks, memory leaks, race conditions, SOLID/DRY violations, and production failures without providing ready-made solutions.

Technical documentation specialist. Creates and maintains the docs/ folder and root README.md. Stack-agnostic.

Harness performance evaluator. Reads all execution traces in docs/harness-history/traces/, computes composite scores per skill_chain, identifies the Pareto frontier of best harness configurations, and recommends the optimal chain for the next session. Run periodically or on demand to guide harness optimization.

Execution trace recorder. Captures what happened during a skill session and persists structured logs to docs/harness-history/traces/. Enables retrospective analysis and harness optimization via harness-evaluator and meta-harness.

Autonomous Meta-Harness proposer. Reads the full harness history filesystem, diagnoses failure patterns, proposes a targeted improvement, stores the candidate, and outputs a JSON decision.