• Agent autonomy level schemes compared
• Agent instruction files compared
• Agent interoperability protocols compared
• AI code review tools compared
• The debt family in AI-assisted development
• Does every agent action need a human?
• Empirical studies of agent-authored pull requests
• Moving agent rules out of the prompt into deterministic enforcement
• Multi-agent orchestration patterns compared
• Organizational adoption of AI coding agents
• Prompt-injection defense patterns compared
• Spec-driven development tools compared
• The SWE-bench family of benchmarks
• What Counts as a Harness
• What does “vibe coding” mean?
• Working within a finite context window