After a year of agentic AI coding, verification—not writing code—became the bottleneck
Positive_Picture-1 · reddit · 2026-11-22
A developer reflects on a year of agentic AI development:
- Early detailed-design-doc-driven delegation failed badly: broken event handling, poor separation of concerns, monolithic code, even hardcoded auth keys; most tasks sank into endless bug-fix cycles.
- As his personal project's architecture matured (and models improved), he documented the whole codebase in Markdown in the repo and added automated tests. Now he just defines requirements and delegates implementation with solid results.
- The new bottleneck: verification. He spends far more time validating AI output, so total dev time hasn't dropped nearly as much as expected.
He asks: can you really delegate the verification process itself to AI?
More from coding & agent
- Garmin Health MCP Server lets you query VO2 max, sleep and training data via natural language — modelcontextprotocol · 2026-11-22
- Stripe MCP connector offers read-only access to Stripe data via API key — modelcontextprotocol · 2026-11-22
- Dev ports Inkling 2 Mini to vLLM, hitting ~200 tok/s decode on 2x RTX 3090 vs 50 on llama.cpp — peaster_ · 2026-11-22
- Six OAuth pitfalls from wiring Claude and ChatGPT into a remote MCP server — Dheeth · 2026-11-22
- Open-source Paperwire MCP lets AI agents send postal mail, certified letters, checks and faxes — QuixiAI · 2026-11-22
- Grok Bot tip: turn repetitive routines into Python scripts to slash usage — nima_owji · 2026-11-22