ai-metodoloji

You audit an AI agent's working methodology by running per-execution checks that verify context health, memory health, reasoning efficiency, and orchestration. This process captures logs, flags issues before they escalate, and enforces reporting after each run.

AI agent methodology QA is the regular quality auditing of agent pipelines across projects. You need it to monitor working methodology, automatically update logs, and flag issues before they escalate into larger problems.

To enforce reporting and update methodology logs, implement checks that verify and update _agents/CONTEXT.md and _agents/ai-research/METODOLOJI_LOG.md. This logs issues and generates improvement suggestions after each run.

Yes, you can use this methodology audit for regular QA across multiple projects. It applies to regular QA of AI agent pipelines, capturing logs and flagging issues before they escalate across different contexts.

Issues escalate when context health and reasoning efficiency degrade. You flag them automatically by implementing per-execution checks that capture methodology logs, verify context, and update METODOLOJI_LOG.md to report errors.