dialectical-autocoder

What problem does it solve?

This Skill enables an adversarial coach-player loop to produce higher-quality code by iterating with fresh context per turn and structured evidence. It helps teams validate implementations against strict requirements while reducing analysis paralysis and manual review time.

Core Features & Use Cases

• Fresh-context iterations: Each turn launches a new agent instance to prevent context pollution and maintain objective evaluation.
• Two-agent collaboration: A Player (implementation) and a Coach (validation) work against the requirements to close gaps faster.
• Evidence-based approvals: Verifies modifications against explicit requirements and captures outcomes for audit.
• Escalation on timeout: Automatically escalates when the max number of turns is reached to avoid endless loops.

Quick Start

Start a session by pointing this Skill at your requirements document (e.g., path/to/requirements.md). The system will run up to 5 turns with fresh context and escalate if not approved.